English中文
PUT AI TO WORK
June 18-21, 2019
Beijing, CN

Artificial Intelligence Conference 2019 Speakers

New speakers are added regularly. Please check back to see the latest updates to the agenda.

Filter

Search Speakers

Ziya Ma is the vice president of architecture, graphics, and software as well as a director of data analytics technologies in system software products at Intel. She’s responsible for optimizing big data solutions on the Intel architecture platform, leading open source efforts in the Apache community, and bringing about optimal big data analytics and AI experiences for customers. Her team works across Intel, the open source community, industry, and academia to further Intel’s leadership in big data analytics. Ziya is a cofounder of the Women in Big Data Forum. At the 2018 Global Women Economic Forum, she was honored as Women of the Decade in Data and Analytics. She holds a master’s degree and PhD in computer science and engineering from Arizona State University.

马子雅是英特尔架构、图形和软件副总裁,以及英特尔系统软件产品部的数据分析技术总监。她主要负责优化英特尔架构平台上的大数据解决方案,领导在Apache社区开展的开源工作,并为客户带来最佳的大数据分析和人工智能体验。她的团队参与了英特尔内部、开源社区、工业界和学术界的工作,以进一步巩固英特尔在大数据分析领域的领导地位。马子雅是“女性在大数据”论坛的联合创始人。在2018年全球妇女经济论坛上,她获得“近10年杰出的数据与分析女性”的荣誉。马子雅在亚利桑那州立大学获得了计算机科学与工程专业的硕士和博士学位。

Presentations

统一大数据分析和人工智能从而更快地大规模洞察(Unifying analytics and AI on big data for faster insights at scale) 主题演讲 (Keynote)

Ziya Ma walks you through Intel’s scalable data insights strategy and related big data analytics and AI technologies such as Analytics Zoo—an end-to-end analytics and AI pipeline for developing full solutions with Apache Spark on Intel Xeon and Intel Optane DC Persistent Memory at scale. She highlights customers use cases and collaboration with industry leaders throughout.

Orchlon Ann is a data engineer at Rakuten in charge of building the Data Science Platform.

Presentations

乐天构建数据科学平台最佳实践(Best practices for building a data science platform at Rakuten) 40分钟议题 (40-minute session)

Orchlon Ann and TzuLin Chin offer an overview of the Data Science Platform, a suite of tools for exploring data, training models, and running GPU/CPU compute jobs in an isolated container environment. Discover it's benefits, including one-click machine learning environment creation, a powerful job scheduler, and a flexible function-as-a-service component.

Mikio Braun is a principal engineer for search at Zalando, one of Europe’s biggest fashion platforms. He worked in research for a number of years before becoming interested in putting research results to good use in the industry. Mikio holds a PhD in machine learning.

Presentations

AI and retail 主题演讲 (Keynote)

What do your customers want? What are the current and upcoming trends? Mikio Braun takes a look at Zalando and the retail industry to explore how AI is redefining the way ecommerce sites interact with customers to create a personalized experience that strives to make sure customers find what they want when they need it.

Architecting AI applications 40分钟议题 (40-minute session)

Mikio Braun takes you through the past 20 years of machine learning research to explore aspects of artificial intelligence, then examines current examples like autonomous cars and chatbots. Together you'll put together a mental model for a reference architecture for artificial intelligence systems.

Chris Butler is the chief product architect at IPsoft. Previously, Chris worked at Microsoft, KAYAK, and Waze, and he was involved in AI-related projects at his startup Complete Seating (data science and constraint programming), Horizon Ventures (advising portfolio companies like Affectiva), and Philosophie (AI consulting and coaching). He was first introduced to AI through graph theory and genetic algorithms while studying computer systems engineering at Boston University. He’s created techniques like empathy mapping for the machine and confusion mapping to create cross-team alignment while building AI products.

Presentations

Design thinking for AI 3小时辅导课 (3-hour Tutorial)

Purpose, a well-defined problem, and trust are important factors to any system, especially those that employ AI. Chris Butler borrows from the principles of design thinking to lead you through exercises that help you create more impactful solutions and better align your team.

Yue Cathy Chang is a business executive recognized for sales, business development, and product marketing in high technology.

Cathy co-founded and is currently the CEO of TutumGene, a technology company that aims to accelerate disease curing by providing solutions for gene therapy and regulation of gene expression. She was most recently with Silicon Valley Data Science, a startup (acquired by Apple) that provided business transformation consulting to enterprises and other organizations using data science- and engineering-based solutions. Prior to that, Cathy was employee #1 hired by the CEO at venture-funded software startup Rocana (acquired by Splunk), where she served as Senior Director of Business Development focusing on building and growing long-term relationships, and notably increased sales leads 2x through building and managing indirect revenue channels.

Prior to Rocana, Cathy held multiple strategic roles at blue chip software enterprise companies as well as startups, including Corporate and Business Development at FeedZai and Datameer; Senior product management, product marketing and sales at Symantec and IBM; and Strategic Sourcing Improvement Consulting at Honeywell.

Cathy holds MS and BS degrees in Electrical and Computer Engineering from Carnegie Mellon University, MBA and MS degrees as a Leaders for Global Operations (LGO) duel-degree fellow from MIT, and two patents for her early work in microprocessor logic design.

Presentations

Artificial intelligence meets genomics: Accelerating understanding of our genetic makeup and the use of genome editing to revolutionize medicine 40分钟议题 (40-minute session)

Genome editing has been dubbed a top technology that could create trillion-dollar markets. Learn how recent advancements in the application of AI to genomic editing are accelerating transformation of medicine with Yue Cathy Chang as she explores how AI is applied to genome sequencing and editing, the potential to correct mutations, and questions on using genome editing to optimize human health.

Dongfeng Chen is the engineering director of Clobotics, a global leader in computer vision solutions for the wind power and retail industries. Clobotics has end-to-end solutions that combine computer vision, artificial intelligence/machine learning, and data analytics software with different hardware form factors, including autonomous drones, mobile applications, and other IoT devices to help companies automate time-intensive operational processes, increase efficiencies, and boost the bottom line using real-time, data-driven, and actionable insights. Clobotics has headquarters in Shanghai and Seattle and has expanded its footprint to Beijing, Dalian, and Singapore. Dongfeng currently leads the Clobotics retail research and development team in Shanghai.

Previously, he was a senior architect at Baidu. He recruited, led, and built a team of more than 30 members, including developers, testers, and product managers, and he created the core algorithms for Baidu advertisement and Baidu Kuaixing (online travel booking site), and Baidu Mall (a flash-sale ecommerce platform). He’s an expert in machine learning and distributed systems. His team developed a way to associate the Baidu search pool with paid advertisements, this in turn brought in more than tens of millions of US dollars in revenue, and this technology is still being used until today. In 2010, his team developed China’s first Groupon website for the travel industry; it was the first travel website that offers group bundle deals. Statistics shows the website accounts for 30%–50% of the market share of short-distance travel near Shanghai. This experience gave him a great foundation to build Baidu Kuaixing, Baidu’s online travel booking site, later on. He received his PhD in computer science from North Carolina State University. His thesis topic was on using structured views to optimize query in information integration.

陈东锋 博士 扩博智能高级研发总监。加入扩博智能之前,陈东锋博士曾担任百度高级架构师。任职期间,陈东锋博士管理和带领研发团 队专注于百度电商广告和百度快行业务的软件研发、测试和产品管理。并为百度快行、特卖频道和电 商知心等等项目开发了核心算法。陈东锋博士的工作成果通过知识图谱技术的创新运用把百度搜索池 与电商广告强相关,为百度在电商特卖领域带来从 0 到上亿人民币的持续巨额营收,该核心技术沿用 至今。陈东锋博士带领团队开发的百度快行项目的成功,使得百度汽车票和火车票交易业务在两年内 成为百度 O2O 垂直行业 GMV第一名,用户体验大大提升。
加入百度之前,陈东锋博士是一名经验丰富的连续创业者。2010 年陈东锋博士带领团队开发了国 内第一个旅游行业团购网站,据数据统计该网站占上海周边短途旅游交易额 30-50%市场份额,这也 为陈东锋博士研发百度出行业务提供了借鉴意义。
陈东锋博士拥有北卡罗来纳州立大学(North Carolina State University)计算机科学博士学位。他 的博士论文主题是使用结构化视图来优化信息集成中的查询。
陈东锋博士是扩博智能智慧零售研发负责人,负责产品研发及交付。扩博智能聚焦计算机视觉和机 器学习技术,专注为行业企业用户提供端到端一体化智能服务,能大力提升传统行业运营效率,加快 数字化变革,所服务的行业包括零售和风电。扩博智能总部位于中国上海和美国西雅图,在北京和大 连设有办事处,新加坡设有分公司。

Presentations

AI如何彻底改变风电行业(How AI is revolutionizing the wind power industry) 40分钟议题 (40-minute session)

One of the biggest challenges to growth remains the high costs of constructing wind farms, as well as the ongoing operations and maintenance costs. Dongfeng Chen dives into the successes and failures of creating an entirely autonomous visual-recognition-powered drone inspection solution for turbine blades, which increased the efficiency by 10 times.

Roger Chen is cofounder and CEO of Computable and program chair for the O’Reilly Artificial Intelligence Conference. Previously, he was a principal at O’Reilly AlphaTech Ventures (OATV), where he invested in and worked with early-stage startups primarily in the realm of data, machine learning, and robotics. Roger has a deep and hands-on history with technology. Before startups and venture capital, he was an engineer at Oracle, EMC, and Vicor. He also developed novel nanoscale and quantum optics technology as a PhD researcher at UC Berkeley. Roger holds a BS from Boston University and a PhD from UC Berkeley, both in electrical engineering.

Presentations

Accelerating AI adoption 主题演讲 (Keynote)

Accelerating AI Adoption

Closing remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason (Jinquan) Dai, and Roger Chen close the first day of keynotes.

Closing remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason (Jinquan) Dai, and Roger Chen close the second day of keynotes.

Decentralized governance of data 40分钟议题 (40-minute session)

Roger Chen details how to enable powerful data lineage properties with decentralized data governance models using blockchain technology. As a result, organizations can easily satisfy growing compliance regulations around data privacy while gaining access to rich external data resources for building AI models.

Friday opening remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason Dai, and Roger Chen open the second day of keynotes.

Thursday opening remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason Dai, and Roger Chen open the first day of keynotes.

Yijing Chen is a senior data scientist in the Cloud AI Group at Microsoft, where she works with external customers in areas such as energy demand forecast, user mobile behavioral analysis, retail demand forecast, energy theft detection, product pricing, and medical claim denial prediction as well as on other projects using various machine learning methods. Yijing holds an MA in statistics from Harvard University.

Presentations

基于深度学习的时间序列预测 (Deep learning for time series forecasting) 3小时辅导课 (3-hour Tutorial)

Almost every business today uses forecasting to make better decisions and allocate its resources more effectively. Deep learning has achieved a lot of success in computer vision, text, and speech processing but has only recently been applied to time series forecasting. Join in to learn how and when to apply deep neural networks to time series forecasting. (presented in Chinese and English)

TzuLin Chin is the data engineer working at Rakuten who originated and lead the team building the Data Science Platform.

Presentations

乐天构建数据科学平台最佳实践(Best practices for building a data science platform at Rakuten) 40分钟议题 (40-minute session)

Orchlon Ann and TzuLin Chin offer an overview of the Data Science Platform, a suite of tools for exploring data, training models, and running GPU/CPU compute jobs in an isolated container environment. Discover it's benefits, including one-click machine learning environment creation, a powerful job scheduler, and a flexible function-as-a-service component.

Dr. Jike Chong is an accomplished executive and professor with experience across industry and academia.

Jike currently heads Data Science, Hiring Marketplace at LinkedIn. He was most recently the chief data scientist at Acorns, the leading micro-investment app in US with over four million verified investors, which uses behavioral economics to help the up-and-coming save and invest for a better financial future. Previously, Jike was the chief data scientist at Yirendai, an online P2P lending platform with more than $7B loans originated and the first of its kind from China to go public on NYSE; established and headed the data science division at Simply Hired, a leading job search engine in Silicon Valley; advised the Obama administration on using AI to reducing unemployment; and led quantitative risk analytics at Silver Lake Kraftwerk, where he was responsible for applying big data techniques to risk analysis of venture investment.

Jike is also an adjunct professor and PhD advisor in the Department of Electrical and Computer Engineering at Carnegie Mellon University, where he established the CUDA Research Center and CUDATeaching Center, which focus on the application of GPUs for machine learning. Recently, he also developed and taught a new graduate level course on machine learning for Internet finance at Tsinghua University in Beijing, China, where he is serving as an adjunct professor.

Jike holds MS and BS degrees in electrical and computer engineering from Carnegie Mellon University and a PhD from the University of California, Berkeley. He holds 11 patents (six granted, five pending).

Presentations

量化互联网金融信用与反欺诈风控 2天培训 (2-day Training)

您想了解金融企业是怎样利用大数据和人工智能技术来画像个人行为并检测欺诈用户的吗?互联网金融幕后的量化分析流程是怎么杨的?个人信用是怎样通过大数据被量化的?在实践过程中,机器学习算法的应用存在着哪些需要关注的方面?怎样通过图谱分析来融合多维数据,为我们区分正常用户和欺诈用户? 这套辅导课基于清华大学交叉信息研究院开设的一门"量化金融信用与风控分析”研究生课。其中会用LendingClub的真实借贷数据做为案例,解说一些具体模型的实现。

崔宏宇,现任DataVisor中国区技术负责人,自2015年起在DataVisor开发使用分布式无监督机器学习算法进行反欺诈检测。负责过如Pinterest、Yelp、阿里巴巴和猎豹移动等大型互联网企业的机器注册、虚假评论、垃圾邮件、欺诈交易和虚假应用安装等场景的反欺诈建模 。在模型调优、特征工程和算法开发等领域都有着丰富的经验。崔宏宇拥有在爱荷华州立大学的博士学位,在博士期间的研究方向为数据分析和结构 – 性能建模等。

Presentations

运用自动化AI技术打击“智能化”网络欺诈 40分钟议题 (40-minute session)

AI技术在赋能各个产业的同时,也被网络黑产所利用,使得黑产攻击更加自动化,更加隐蔽,难于检测。 DataVisor在互联网反欺诈领域研究发现,目前黑产的攻击模型呈现以下趋势:攻击方法多样化而变化快,攻击手段趋于模拟正常用户,攻击账号主要来源由大规模注册渐渐转向ATO账号。传统的规则系统和有监督的模型,由于对欺诈案例以及标签数据的强依赖,往往无法及时应对迅速演化的黑产攻击,在反欺诈中一直处于被动防守的状态。DataVisor的无监督算法,通过全局分析,在高维空间聚类,可以在无标签情况下,自动发现大规模关联欺诈团伙。无监督算法在提前预警以及检测快速演变欺诈模式方面体现了显著的优势。

Jason (Jinquan) Dai is a senior principal engineer and CTO of big data technologies at Intel, where he is responsible for leading the global engineering teams (located in both Silicon Valley and Shanghai) on the development of advanced big data analytics (including distributed machine and deep learning), as well as collaborations with leading research labs (e.g., UC Berkeley AMPLab and RISELab). Jason is an internationally recognized expert on big data, cloud, and distributed machine learning; he is the program cochair of the O’Reilly AI Conference in Beijing, a founding committer and PMC member of Apache Spark, and the creator of BigDL, a distributed deep learning framework on Apache Spark.

Presentations

Closing remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason (Jinquan) Dai, and Roger Chen close the first day of keynotes.

Closing remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason (Jinquan) Dai, and Roger Chen close the second day of keynotes.

Friday opening remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason Dai, and Roger Chen open the second day of keynotes.

Thursday opening remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason Dai, and Roger Chen open the first day of keynotes.

Bin Fan is a software engineer at Alluxio and a PMC member of the Alluxio project. Previously, Bin worked at Google, building next-generation storage infrastructure, where he won Google’s technical infrastructure award. He holds a PhD in computer science from Carnegie Mellon University.

Presentations

Atom:Supremind云原生深度学习平台(Atom:A cloud native deep learning platform at Supremind) 40分钟议题 (40-minute session)

The Atlab Lab at Qiniu Cloud focuses on deep learning for computer vision. Chaoguang Li, Haoyuan Li, and Bin Fan lead a deep dive into AVA, a high-performance and cost-effective cloud-based training platform for deep learning, which deeply integrates an open source software stack including TensorFlow, Caffe, Alluxio, and KODO, the company's own cloud object storage.

Xing Fan is CTO at Squirrel AI. He leads the research and development of technologies and products in the IM and SNS areas and has deep understanding of the relevant high-concurrency tasks, web, and various client research and development. Xing has decades of experience in internet technology architecture and management. He also has a deep experience in technical team management and rapid team growth for startups. Previously, he was the CTO and R&D director at IMO Cloud Office, the department manager at Shanda Networks, the server manager at 51.com, and the senior engineer at Tencent.

Presentations

Squirrel AI Learning的AI导师:AI-adaptive技术在K-12教育中的实际应用(Squirrel AI Learning’s AI tutors: Real-life applications of AI-adaptive technology in K–12 education) 40分钟议题 (40-minute session)

Squirrel AI Learning is the first artificial intelligence technology company in China to apply AI-adaptive technology to K–12 education. Xing Fan dives deep into its implementation approach and teaches you about the business process, pedagogy, architecture, operation, and theoretical underpinning of this adaptive learning service.

中国人寿研发中心高级工程师,自2014年从事大数据相关项目开发及管理。2016年开始研究机器学习模型的构建与实施,已主导多个模型落地实施。

Presentations

保险中的机器学习实践 40分钟议题 (40-minute session)

分析保险行业人工智能发展情况及现有数据特性,评估机器学习模型构建的主流工具、语言、算法。总结基于机器学习技术,实现一个保险业人工智能场景的全流程——从场景研讨、数据加工提取到模型构建、模型效果评估、模型落地实施。以一个真实的机器学习模型项目为例,介绍整个方法论不同环节中各方人员的参与工作内容和比例,探讨特征稳定性、样本不均衡、参数选择、模型可解释性等环节的难点及尝试方案。为金融或者其他行业的机器学习项目落地提供参考和指导。

Huaxin Gao is a software engineer in IBM’s Open Source Data and AI Group focusing on Apache Spark machine learning and deep learning. She’s an active code contributor to the Apache Spark project.

Presentations

AI pipelines on container platforms 40分钟议题 (40-minute session)

AI pipelines simplify the lifecycle workflow management and enhance reproducibility and collaboration for machine learning and deep learning projects. Cloud native platform solutions offer great portability and scalability. Weiqiang Zhuang and Huaxin Gao show how, by combining strengths, AI pipelines on container platforms can help accelerate AI application development and deployment.

Bas Geerdink is a programmer, scientist, and IT manager at ING, where he’s responsible for the fast data systems that process and analyze streaming data. Bas has a background in software development, design, and architecture with broad technical experience from C++ to Prolog to Scala. His academic background is in artificial intelligence and informatics. Bas’s research on reference architectures for big data solutions was published at the IEEE conference ICITST 2013. He occasionally teaches programming courses and is a regular speaker at conferences and informal meetings.

Presentations

AI at ING: The why, how, and what of a data-driven enterprise 40分钟议题 (40-minute session)

ING is a data-driven enterprise, with analytics skills as a top strategic priority. AI is at the core of ING’s business, and the company is investing in AI, big data, and analytics to improve business processes such as balance forecasting, fraud detection, and customer relation management. Follow along with (and be inspired by) Bas Geerdink's overview of the company's use cases and technology.

Chenhui Hu is a data scientist in the Cloud and AI Division at Microsoft. His current interests include retail forecast, inventory optimization, IoT data, and deep learning. He also has research experience in wireless networks and network data analysis. He’s a recipient of the third IEEE ComSoc Asia-Pacific Outstanding Paper Award. He holds a PhD from Harvard University, where his PhD thesis focused on biomedical imaging data mining.

Presentations

Forecasting customer activities with dilated convolution neural networks: Use case and best practices 40分钟议题 (40-minute session)

Forecasting customer activities is an important, common business problem, and Tao Lu and Chenhui Hu forecast customer behavior based on billions of user activities. Join them as they share how Microsoft improved forecasting accuracy by 25% with dilated convolutional neural networks and reduced time in development by 80% with a set of time series forecasting best practices.

Klein Hu is the senior software engineer on the Microsoft Azure machine learning team, focusing on the AI model inferencing area, especially ONNX model operationalization and acceleration with ONNX Runtime. Klein holds an MS in computer science from Beijing Normal University.

Presentations

ONNX:开放和互操作平台让AI无处不在(AI everywhere: Open and interoperable platform for AI with ONNX) 40分钟议题 (40-minute session)

An open and interoperable ecosystem enables you to choose the framework that's right for you, train at scale, and deploy to cloud and edge. ONNX provides a common format supported by many popular frameworks and hardware accelerators. Henry Zeng, Klein Hu, and Emma Ning introduce you to ONNX and its core concepts. (Presented in English and Chinese.)

Kai Huang is a software engineer at Intel. His work mainly focuses on developing deep learning frameworks on Apache Spark and helping customers work out end to end deep learning solutions on big data platforms. He is a main contributor to Analytics Zoo and BigDL.

Kai Huang是英特尔软件工程师。 他的工作主要集中在开发Apache Spark深度学习框架,并帮助客户在大数据平台上制定端到端深度学习解决方案。他是Analytics Zoo和BigDL的主要贡献者之一。

Presentations

Analytics Zoo:基于Apache Spark的分布式TensorFlow和Keras(Analytics Zoo: Distributed TensorFlow and Keras on Apache Spark) 3小时辅导课 (3-hour Tutorial)

Zhichao Li, Kai Huang, and Yang Wang show you how to build and productionize deep learning applications for big data using Analytics Zoo—a unified analytics and AI platform that seamlessly unites Spark, TensorFlow, Keras, and BigDL programs into an integrated pipeline—illustrated though real-world use cases from JD.com, MLSListings, the World Bank, Baosight, and Midea/KUKA.

Office Depot利用基于Apache Spark的深度学习实现实时产品推荐(Real-time product recommendations leveraging deep learning on Apache Spark in Office Depot) 40分钟议题 (40-minute session)

Real-time recommender systems are critical for the success of the ecommerce industry. Join Kai Huang, Luyang Wang, and Jing Kong as they showcase how to build efficient recommender systems for the ecommerce industry using deep learning technologies.

Alex Ingerman is a product manager at Google AI, focusing on federated learning and other privacy-preserving technologies. His mission is to enable all ML practitioners to protect their users’ privacy by default. Previously, Alex worked on ML-as-a-service platforms for developers, web-scale search, content recommendation systems, and immersive data exploration and visualization. Alex lives in Seattle, where as a frequent bike and occasional kayak commuter, he has fully embraced the rain. Alex holds a BS in computer science and an MS in medical engineering.

Presentations

The future of machine learning is decentralized 40分钟议题 (40-minute session)

Federated learning involves training ML models across a fleet of participating devices without collecting their data in a central location. Alex Ingerman examines federated learning, compares the traditional and federated ML workflows, and explores the current and upcoming use cases for decentralized machine learning with examples from Google's deployment of this technology.

Jewel James is a data scientist at Gojek.

Presentations

Using ML for personalizing food recommendations 40分钟议题 (40-minute session)

Hear how Maulik Soneji and Jewel James prototyped the search framework that personalizes the restaurant search results by using machine learning (ML) to learn what constitutes a relevant restaurant given a user's purchasing history.

Michael James is a pioneer in geometrically mapped algorithms and the founder and chief architect in advanced technologies such as mathematics, algorithms and software at Cerebras Systems—a computer hardware company developing deep technologies to scale and accelerate machine learning by orders of magnitude for AI applications. Previously, Michael was a fellow at Advanced Micro Devices (AMD), where under his leadership, the team designed first-of-its-kind technology based on a self-healing fabric interconnect to allow reliable operation of large computer clusters, and was chief architect at SeaMicro Systems, where he specialized in real-time workload placement and routing algorithms. His experience includes the fields of computer-automated language translation, algorithms for gesture recognition, compilers, operating systems, and microcontroller design. Michael also provides advice and gives talks on a diverse range of AI topics to established Silicon Valley companies. His passion for AI comes from his roots in academia: he holds bachelor’s degrees from UC Berkeley in mathematics, computer science, and neurobiology.

Presentations

Designing computer hardware for artificial intelligence 主题演讲 (Keynote)

Artificial intelligence is defining a new generation of computer technology with applications that blur the boundaries between intuition, art, and science. Michael James examines the fundamental drivers of computer technology, surveys the landscape of AI hardware solutions, and explores the limits of what's possible as new computer platforms emerge.

Yangqing Jia leads Alibaba’s AI and Big Data org, supporting the large-scale applications both inside the company and on Aliyun, the number one cloud provider in China and a market leader globally. The org provides advanced AI systems and service combined with conventional big data wisdom (EMR, Flink, and Spark) as well as battle-tested solutions to serve every cloud client.

Presentations

为什么说人工智能和云计算乃天作之合?(Why do we say AI Should be Cloud Native?) 主题演讲 (Keynote)

The recent years of AI has grown out of labs and created a transformative power for a vast range of industries. But, while we take it for granted that AI and Cloud come hand in hand, I'll show you an argument one step further: AI should be Cloud Native.

Tim Kraska is an associate professor of electrical engineering and computer science in MIT’s Computer Science and Artificial Intelligence Laboratory and codirector of the Data System and AI Lab at MIT (DSAIL@CSAIL). His research focuses on building systems for machine learning and using machine learning for systems. Previously, Tim was an assistant professor at Brown, spent time at Google Brain, and was a postdoc in the AMPLab at UC Berkeley after his PhD at ETH Zurich. Tim’s a 2017 Alfred P. Sloan Research Fellow in computer science and received several awards including the 2018 VLDB Early Career Research Contribution Award, the 2017 VMware Systems Research Award, an NSF CAREER Award, as well as several best paper and demo awards at VLDB and ICDE.

Presentations

Toward learned algorithms, data structures, and systems 主题演讲 (Keynote)

Systems and applications are composed from basic data structures and algorithms. Most of these have been around since the beginnings of CS and form every intro lecture. Yet, we might soon face an inflection point. Tim Kraska outlines different ways to build learned algorithms and data structures to achieve instance optimality and unprecedented performance for a wide range of applications.

Chenta Lee is a senior software engineer at IBM Security Systems, where he’s the architect of IBM Security Network Protection and currently focuses on the network security in the cloud. His expertise includes emerging cloud technologies, with seven years of experience in cloud security products, as well as software-defined networking, virtualization and advanced threat protection.

Presentations

Detect the undetectable at the breach 40分钟议题 (40-minute session)

By combining various analytics including DGA, squatting, tunneling, and rebinding detection, it's possible to build a DNS analytic playbook to anneal actionable threat intelligence from billions of DNS requests. Chenta Lee outlines how DNS volumetric data and analytics complement each other to create a new dimension to look at security postures and how to leverage it in security operations.

Chaoguang Li is the senior architect and director of AI at Qiniu Cloud. He’s been working in distributed systems for more than 10 years. Previously, he worked on the first generation of SSD tiered storage DS8000 at IBM and was the chief architect of the all-flash storage Dorado Cache in Huawei.

Presentations

Atom:Supremind云原生深度学习平台(Atom:A cloud native deep learning platform at Supremind) 40分钟议题 (40-minute session)

The Atlab Lab at Qiniu Cloud focuses on deep learning for computer vision. Chaoguang Li, Haoyuan Li, and Bin Fan lead a deep dive into AVA, a high-performance and cost-effective cloud-based training platform for deep learning, which deeply integrates an open source software stack including TensorFlow, Caffe, Alluxio, and KODO, the company's own cloud object storage.

Haoyuan (H.Y.) Li is the founder, chairman, and CTO of Alluxio. He holds a PhD in computer science from UC Berkeley’s AMPLab, where he created the Alluxio (formerly Tachyon) open source data orchestration system, cocreated Apache Spark Streaming, and became an Apache Spark founding committer. He also holds an MS from Cornell University and a BS from Peking University, both in computer science.

Presentations

Atom:Supremind云原生深度学习平台(Atom:A cloud native deep learning platform at Supremind) 40分钟议题 (40-minute session)

The Atlab Lab at Qiniu Cloud focuses on deep learning for computer vision. Chaoguang Li, Haoyuan Li, and Bin Fan lead a deep dive into AVA, a high-performance and cost-effective cloud-based training platform for deep learning, which deeply integrates an open source software stack including TensorFlow, Caffe, Alluxio, and KODO, the company's own cloud object storage.

Data orchestration for AI, big data, and the cloud 主题演讲 (Keynote)

Haoyuan Li offers an overview of a data orchestration layer that provides a unified data access and caching layer for single cloud, hybrid, and multicloud deployments. It enables distributed compute engines like Presto, TensorFlow, and PyTorch to transparently access data from various storage systems while actively leveraging an in-memory cache to accelerate data access.

Renjie Li is the head of NetEase Fuxi Lab, the first AI research institute focusing on games in China. Previously, Renjie was the head of data (data architecture, data science, and artificial intelligence) at Riot Games, where he built the Data Department from scratch. He holds a PhD from the University of Rochester and a BS from the Special Class for Gifted Young at the University of Science and Technology of China.

Presentations

Enlighten the future of games with artificial intelligence 40分钟议题 (40-minute session)

Theoretical AI research isn't a bottleneck in AI, but finding a good application scenario for AI is. Renjei Li explains why gaming is a great scenario for AI and walks you through recent research in the future of AI games involving reinforcement learning, natural language processing (NLP), computer vision and graphics, and user personas and virtual humans.

Zhichao Li is a senior software engineer at Intel focused on distributed machine learning, especially large-scale analytical applications and infrastructure on Spark. He’s also an active contributor to Spark. Previously, Zhichao worked in Morgan Stanley’s FX Department.

Zhichao Li是英特尔的高级软件工程师,专注于分布式机器学习,尤其是Spark上的大规模分析应用和基础架构。他也是一名Spark项目的积极贡献者。 在加入英特尔之前,Zhichao曾在摩根士丹利的外汇部工作。

Presentations

Analytics Zoo:基于Apache Spark的分布式TensorFlow和Keras(Analytics Zoo: Distributed TensorFlow and Keras on Apache Spark) 3小时辅导课 (3-hour Tutorial)

Zhichao Li, Kai Huang, and Yang Wang show you how to build and productionize deep learning applications for big data using Analytics Zoo—a unified analytics and AI platform that seamlessly unites Spark, TensorFlow, Keras, and BigDL programs into an integrated pipeline—illustrated though real-world use cases from JD.com, MLSListings, the World Bank, Baosight, and Midea/KUKA.

Jianxun Lian is now an associate researcher at Microsoft Research Asia. He’s published research papers as the first author on top international conferences such as KDD, WWW, and IJCAI. He received his PhD from the University of Science and Technology of China. His research interests include deep learning-based recommender systems and user modeling. In addition to academic research, he likes building practical systems and he was among the top winning teams in several data contests including IJCAI 2015, CIKM 2016, and RecSys Challenge 2017.

Presentations

Best practices for building enterprise-grade recommendation systems on Azure with Microsoft/Recommenders 40分钟议题 (40-minute session)

Enterprises benefit from recommendation systems for revenue and customer engagement, but creating such a system is time-consuming. Le Zhang and Jianxun Lian explore the Microsoft/Recommenders repository, which offers solutions to building recommendation systems. It contains classic and state-of-the-art algorithms from Microsoft and enables enterprise success by leveraging Azure's cloud capability.

Richard Liaw is a PhD student in the Berkeley Artificial Intelligence Research (BAIR) Lab and RISELab at UC Berkeley working with Joseph Gonzalez, Ion Stoica, and Ken Goldberg. He’s worked on a variety of different areas, ranging from robotics to reinforcement learning to distributed systems. He’s working on Ray, a distributed execution engine for AI applications; RLlib, a scalable reinforcement learning library; and Tune, a distributed framework for model training.

Presentations

Building reinforcement learning models and AI applications with Ray 3小时辅导课 (3-hour Tutorial)

Ray is a general-purpose framework for programming your cluster. Richard Liaw leads a deep dive into Ray, walking you through its API and system architecture and sharing application examples, including several state-of-the-art AI algorithms.

刘影,现任鲸算科技高级风控经理,集团科技教育品牌鲸小小联合创始人。武汉大学学士,加拿大纽芬兰纪念大学硕士,电子工程专业,海外读研期间当过研究员,曾在雷达信号处理领域发表过两篇SCI,参加过三次国际学术会议。2014年底毕业后在青岛墨尔文中学担任一年全英文数学老师,教L5-U6的中学生IGCSE/ALEVEL数学课,并组织了青少年编程俱乐部。2016年加入鲸算科技(原闪银Wecash),任职数据科学家一年,从事互联网金融信用评估特征工程搭建和线上模型研发。2017年至今,转岗为风控高级经理,致力于公司的数据管理工作,深度挖掘数据商业价值,与资方、财务、产品运营、催收、人事管理等团队紧密合作,设计AB实验,帮助大家为公司降本增效,希望把数据科学技术落地为更大更广的商业价值。

像所有其他从事数据科学的同行一样,渴望看见AI技术在更多领域落地,产生影响力。也想鼓励更多女性同行加入到这项激动人心的事业中来。在设计AI产品时,像”Design of Everyday Things”书中讲到的,在追求理性效率同时,倡导更多人文关怀,人性化地去解决问题,共同提高AI落地的方法论,造福人类未来生活。

Ying (Claire) Liu is a senior risk management manager at Abakus and cofounder of Abakus Kids (a new edtech brand founded in 2018). As a research student, she published two SCI papers and attended several international conferences in radar signal processing in 2012–2014. Previously in China, she taught L5–U6 kids IGCSE/ALEVEL math classes and held a web development club at Malvern College Qingdao. She was a data scientist at Abakus Group (Wecash China) working on everything from feature engineering to model deploy, witnessing AI technology accelerate online lending in China. Both a data officer and product manager, she dove into the architecture of the data management platform, shared the knowledge of data in the company, and spared no effort to help people do better on creating business value, collaborating closely with product, operation, finance, and collection teams with data. She holds a bachelor of engineering in radio physics (radio wave propagation and antenna) from Wuhan University and a master of engineering in electrical and computer engineering from Memorial University of Newfoundland in Canada. She can be found on LinkedIn.

Presentations

一个改善债务催收的AI解决方案(A humane AI solution to improve debt collection) 40分钟议题 (40-minute session)

Abakus's AI debt collection platform provides a friendly and humane product solution designed for people who work on the frontline: live agents of the organization. The company's agent training has been enhanced with an AI-friendly culture. Join Ying Liu as she details the results of an experiment showing how the company improved the performance of the collection assistants.

Ben Lorica is the chief data scientist at O’Reilly Media. Ben has applied business intelligence, data mining, machine learning, and statistical analysis in a variety of settings, including direct marketing, consumer and market research, targeted advertising, text mining, and financial engineering. His background includes stints with an investment management company, internet startups, and financial services.

Presentations

Accelerating AI adoption 主题演讲 (Keynote)

Accelerating AI Adoption

Closing remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason (Jinquan) Dai, and Roger Chen close the first day of keynotes.

Closing remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason (Jinquan) Dai, and Roger Chen close the second day of keynotes.

Friday opening remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason Dai, and Roger Chen open the second day of keynotes.

Thursday opening remarks 主题演讲 (Keynote)

Program chairs Ben Lorica, Jason Dai, and Roger Chen open the first day of keynotes.

David Low is the cofounder and chief data scientist at Pand.ai, a company building an AI-powered chatbot to disrupt and shape the booming conversational commerce space with deep natural language processing. He represented Singapore and the National University of Singapore (NUS) in the 2016 Data Science Games held in France, and clinched the top spot among Asian and American teams. David has been invited as a guest lecturer by NUS to conduct master classes on applied machine learning and deep learning topics. Throughout his career, David has engaged in data science projects across manufacturing, telco, ecommerce, and the insurance industry, including sales forecast modeling and influencer detection, which won him awards in several competitions and was featured on the IDA website and the NUS publication. Previously, he was a data scientist at the Infocomm Development Authority (IDA) of Singapore and was involved in research collaborations with Carnegie Mellon University (CMU) and Massachusetts Institute of Technology (MIT) on projects funded by the National Research Foundation and SMART. He competes on Kaggle and holds a top 0.2% worldwide ranking.

Presentations

The unreasonable effectiveness of transfer learning on natural language processing 40分钟议题 (40-minute session)

Transfer learning has been a tremendous success in computer vision as a result of the ImageNet competition. In the past few months, natural language processing (NLP) has witnessed several breakthroughs with transfer learning, namely ELMo, Transformer, ULMFit, and BERT. Join David Low as he showcases the use of transfer learning on NLP applications with state-of-the-art accuracy.

Tao Lu is a data scientist in the cloud and AI organization at Microsoft. He has strong background in applying machine learning and deep learning techniques to forecasting problems and deep domain knowledge in cloud identity and the financial services industry. He graduated from University of Washington with a master’s degree in computational finance.

Presentations

Forecasting customer activities with dilated convolution neural networks: Use case and best practices 40分钟议题 (40-minute session)

Forecasting customer activities is an important, common business problem, and Tao Lu and Chenhui Hu forecast customer behavior based on billions of user activities. Join them as they share how Microsoft improved forecasting accuracy by 25% with dilated convolutional neural networks and reduced time in development by 80% with a set of time series forecasting best practices.

Zhenxiao Luo is leading Interactive Query Engines team at Twitter, where he focuses on Druid, Presto, and Hive. Before joining Twitter, Zhenxiao was running Interactive Analytics team at Uber. Previously, he worked on big data related projects at Netflix, Facebook, Cloudera, and Vertica. Zhenxiao is PrestoDB committer. He holds a master’s degree from the University of Wisconsin-Madison and a bachelor’s degree from Fudan University.

Presentations

查询地球:Uber的地理空间大数据分析(Query the planet: Geospatial big data analytics at Uber) 40分钟议题 (40-minute session)

Locations and trips provide insights that can improve business decisions and better serve users, but geospatial data analysis is particularly challenging. It requires efficiency, usability, and scalability in order to meet user needs and business requirements. Join Zhenxiao Luo to learn how Uber uses artificial intelligence to analyze geospatial big data, one of its distinct challenges.

David Maman is founder, CEO, and CTO at Binah. A serial entrepreneur, David founded HexaTier/GreenSQL (acquired by Huawei), Precos, Vanadium-soft, GreenCloud, Teridion, Terrasic, and ReSec, among others. Previously, he was a director in Fortinet’s CTO office, where he managed information security at the Israeli telecom Bezeq. He has 24 years’ experience in leadership, AI, cybersecurity, development, and networking and is a veteran of an elite Israel Defense Forces (IDF) unit. He was named one of the top 40 Israeli internet startup professionals by TheMarker Magazine and one of the top 40 under 40 most promising Israeli business professionals by Globes magazine. David holds a master’s degree in computer science from Open University.

Presentations

Hacking humans made easy: Signal processing + AI + video 40分钟议题 (40-minute session)

Zero-day attacks. IoT-based botnets. Cybercriminal AI versus cyberdefender AI. While these won’t be going away, they aren’t our biggest worry in cybercrime. Hacking humans is. David Maman demonstrates how the combination of minutes of video, signal processing, remote heart-rate monitoring, AI, ML, and data science can identify a person’s health vulnerabilities, which evildoers can make worse.

Eitan Medina is the chief business officer at Habana. Previously, he was the president and general manager of the Fingerprint Business Unit at InvenSense, a TDK Group Company; vice president of marketing at InvenSense; vice president of engineering at Audience, Inc. (acquired by Knowles); vice president of cellular engineering at Marvell; and chief technology officer at Galileo (Acquired by Marvell). He holds a BSc in electrical engineering from the Technion, Israel.

Presentations

A fresh approach to building high-performance AI systems (sponsored by Habana Labs) 40分钟议题 (40-minute session)

The new class of purpose-built AI processors presents data center engineers and developers with opportunities to deliver tangible advancements in AI productivity and efficiency, resulting in lower total cost of ownership. Eitan Medina reveals the advantages derived from new approaches to building high-performance AI systems.

Increasing AI productivity and efficiency with purpose-built AI processors (sponsored by Habana Labs) 主题演讲 (Keynote)

Eitan Medina details advances made possible with AI processors designed to address AI-specific computing requirements, chief among them increasing AI throughput speeds while lowering power consumption. This new class of AI processing brings significantly improved productivity and efficiency to the data center to overcome limitations of existing CPU- and GPU-based solutions.

凡普数据中心算法工程师,NLP&大规模算法技术负责人,负责自然语言处理、语音识别、爬虫相关工作。带领团队自研语音数据平台,在客服、电销等业务中使用。

Presentations

深度学习语音技术在金融场景中的应用 40分钟议题 (40-minute session)

该议题主要包括:1. 语音切分技术的原理和应用;2. 语音识别模型的构建优化;3. 语音情感分析构建应用;4. 语音数据的实时处理框架;5. 金融场景业务落地。

Emma Ning is senior program manager on the Microsoft cloud and AI ML platform team, focusing on AI model operationalization and acceleration with ONNX/ONNXRuntime in support of Microsoft’s strategic investment for open and interoperable AI. She’s been driving search engine experience for more than five years and spent two years exploring adoption of AI among various businesses. Emma holds a MS in computer science from the Institute of Computing Technology, Chinese Academy of Sciences.

Presentations

ONNX:开放和互操作平台让AI无处不在(AI everywhere: Open and interoperable platform for AI with ONNX) 40分钟议题 (40-minute session)

An open and interoperable ecosystem enables you to choose the framework that's right for you, train at scale, and deploy to cloud and edge. ONNX provides a common format supported by many popular frameworks and hardware accelerators. Henry Zeng, Klein Hu, and Emma Ning introduce you to ONNX and its core concepts. (Presented in English and Chinese.)

Richard Ott is a data scientist in residence at the Data Incubator, where he combines his interest in data with his love of teaching. Previously, he was a data scientist and software engineer at Verizon. Rich holds a PhD in particle physics from the Massachusetts Institute of Technology, which he followed with postdoctoral research at the University of California, Davis.

Presentations

Deep learning with PyTorch 2天培训 (2-day Training)

PyTorch is a machine learning library for Python that allows users to build deep neural networks with great flexibility. Rich Ott introduces you to the PyTorch workflow and explores how its easy-to-use API and seamless use of GPUs makes it a sought-after tool for deep learning. Join in to get the knowledge you need to build deep learning models using real-world datasets.

Vanja Paunic is a data scientist in the Algorithms and Data Science Group at Microsoft London. She works on building machine learning solutions with external companies utilizing Microsoft’s AI Cloud Platform. She holds a PhD in computer science with a focus on data mining in the biomedical domain from the University of Minnesota.

Presentations

基于深度学习的时间序列预测 (Deep learning for time series forecasting) 3小时辅导课 (3-hour Tutorial)

Almost every business today uses forecasting to make better decisions and allocate its resources more effectively. Deep learning has achieved a lot of success in computer vision, text, and speech processing but has only recently been applied to time series forecasting. Join in to learn how and when to apply deep neural networks to time series forecasting. (presented in Chinese and English)

Dmitry Pechyoni is a senior data scientist in the Cloud AI Group at Microsoft, where he works on building end-to-end data science solutions in various domains, including retail, energy management, and predictive maintenance. Previously, he built machine learning models for display advertising Akamai and MediaMath. Dmitry holds a PhD in theoretical machine learning from the Technion – Israel Institute of Technology.

Presentations

基于深度学习的时间序列预测 (Deep learning for time series forecasting) 3小时辅导课 (3-hour Tutorial)

Almost every business today uses forecasting to make better decisions and allocate its resources more effectively. Deep learning has achieved a lot of success in computer vision, text, and speech processing but has only recently been applied to time series forecasting. Join in to learn how and when to apply deep neural networks to time series forecasting. (presented in Chinese and English)

Mark Ryan is a leader in the machine learning hub at IBM, where he’s responsible for shepherding customers to a variety of database products, including IBM Integrated Analytics System, which includes a full-blown machine learning environment: DSX. Ever since doing a masters at the University of Toronto in the ‘80s, he’s had an interest in machine learning and artificial intelligence, with a particular focus on deep learning on structured data and natural language processing (NLP).

Presentations

Using deep learning and time series forecasting to reduce transit delays 40分钟议题 (40-minute session)

Toronto is unique among North American cities for having a legacy streetcar network as an integral part of its transit system. This means streetcar delays are a major contributor to gridlock in the city. Learn about applying deep learning time series forecasting to machine learning as Mark Ryan and Alina Li Zhang explain how streetcar delays can be predicted...and prevented.

Kaz Sato is a staff developer advocate on the cloud platform team at Google, where he leads the developer advocacy team for machine learning and data analytics products such as TensorFlow, the Vision API, and BigQuery. Kaz has been leading and supporting developer communities for Google Cloud for over seven years. He’s a frequent speaker at conferences, including Google I/O 2016, Hadoop Summit 2016 San Jose, Strata and Hadoop World 2016, and Google Next 2015 NYC and Tel Aviv, and he has hosted FPGA meetups since 2013.

Presentations

ML ops and Kubeflow pipelines 40分钟议题 (40-minute session)

Kaz Sato explains how creating an ML model is just a starting point. To bring the technology into production service, you need to solve various real-world issues such as building a data pipeline for continuous training, automated validation of the model, version control of the model, scalable serving infra, and ongoing operation of the ML infra with monitoring and alerting.

Alejandro Saucedo is the chief scientist at the Institute for Ethical AI & Machine Learning. In his more than 10 years of software development experience, Alejandro has held technical leadership positions across hypergrowth scale-ups and tech giants including Eigen Technologies, Bloomberg LP, and Hack Partners. Alejandro has a strong track record of building multiple departments of machine learning engineers from scratch and leading the delivery of numerous large-scale machine learning systems across the financial, insurance, legal, transport, manufacturing, and construction sectors in Europe, the US, and Latin America.

Presentations

A practical guide toward explainability and bias evaluation in machine learning 3小时辅导课 (3-hour Tutorial)

Numerous high-profile incidents have proved undesired bias in machine learning a worrying topic. Alejandro Saucedo uses a hands-on example to demystify machine learning bias. You'll automate the loan-approval process for a company and explore key tools and techniques from the latest research that allows you to assess and mitigate undesired bias in machine learning models.

Min Shen is an engineer on LinkedIn’s Hadoop infrastructure development team, helping to build next-generation Hadoop infrastructure at LinkedIn with better performance and manageability. Min holds a PhD in computer science from the University of Illinois, where he focused on distributed computing.

Presentations

领英基于Spark和TensorFlow的大规模AI基础架构 40分钟议题 (40-minute session)

领英公司的几乎所有产品都离不开基于海量数据和大规模数据运算的机器学习模型。怎样构建一个稳定,高效,和易用的人工智能基础架构,越来越成为一个核心的问题。 这个演讲会先介绍领英大数据团队是怎样在5年的时间里演进这个基础架构,从开始的完全基于Spark的系统,到现在Spark+TensorFlow的环境。 我们还会重点介绍近期解决的技术挑战,来应对接近500PB数据和将近6亿会员的巨大经济图谱。这些挑战包括大规模重量级的深度学习模型,Spark的调优,以及在机器学习生产线中连接不同的步骤(数据准备,模型构建,模型训练,在线inference)。 最后我们会介绍我们近期一些成功的深度学习案例,以及团队在AI基础架构上未来2~3年的计划和愿景。

Maulik Soneji is a data engineer at GO-JEK, where he works with different parts of data pipelines for a hypergrowth startup. Outside of learning about mature data systems, he’s interested in Elasticsearch, Go, and Kubernetes.

Presentations

Using ML for personalizing food recommendations 40分钟议题 (40-minute session)

Hear how Maulik Soneji and Jewel James prototyped the search framework that personalizes the restaurant search results by using machine learning (ML) to learn what constitutes a relevant restaurant given a user's purchasing history.

Joseph Spisak is the product manager for Facebook’s AI open source platform, including PyTorch and ONNX. Previously, he led AI partnerships and deep learning products at Amazon Web Services, where he and his team were dedicated to building tools and solutions to help democratize deep learning for the developer community and ultimately accelerate the development of deep learning-based applications. Joseph holds a bachelor’s degree in electrical engineering from Michigan State University and an MBA and MS in finance from the University of Denver. He’s a proud graduate of the Entrepreneurial and Innovation Program at Stanford University Graduate School of Business.

Presentations

Bringing research and production together with PyTorch 1.0 40分钟议题 (40-minute session)

Learn how PyTorch 1.0 enables you to take state-of-the-art research and deploy it quickly at scale in areas from autonomous vehicles to medical imaging. Joseph Spisak dives deep on the latest updates to the PyTorch framework including TorchScript and the JIT compiler, deployment support, and the C++ interface, and explains how Facebook uses PyTorch 1.0 to power AI across its products.

Ion Stoica is a professor in the electrical engineering and computer sciences (EECS) department at the University of California, Berkeley, where he does research on cloud computing and networked computer systems. Previously, he worked on dynamic packet state, chord DHT, internet indirection infrastructure (i3), declarative networks, and large-scale systems, including Apache Spark, Apache Mesos, and Alluxio. He’s the cofounder of Databricks—a startup to commercialize Apache Spark—and Conviva—a startup to commercialize technologies for large-scale video distribution. Ion is an ACM fellow and has received numerous awards, including inclusion in the SIGOPS Hall of Fame (2015), the SIGCOMM Test of Time Award (2011), and the ACM doctoral dissertation award (2001).

Presentations

AI and systems at RISELab 主题演讲 (Keynote)

Ion Stoica outlines a few projects at the intersection of AI and systems that RISELab, at the University of California, Berkeley, is developing. RISELab is the successor of AMPLab, where several highly successful open source projects, including Apache Spark and Apache Mesos, were developed.

Angus Taylor is a data scientist in the Cloud AI Group at Microsoft, where he builds data science solutions for external customers in the retail, energy, engineering, and package distribution sectors. He holds an MSc in AI from the University of Edinburgh.

Presentations

基于深度学习的时间序列预测 (Deep learning for time series forecasting) 3小时辅导课 (3-hour Tutorial)

Almost every business today uses forecasting to make better decisions and allocate its resources more effectively. Deep learning has achieved a lot of success in computer vision, text, and speech processing but has only recently been applied to time series forecasting. Join in to learn how and when to apply deep neural networks to time series forecasting. (presented in Chinese and English)

Arun Verma is the head of the quantitative research solutions team at Bloomberg. He also serves on the board of a nonprofit that helps with humanitarian projects in India serving impoverished children and women in the areas of education and vocational training. Since he joined the Bloomberg Quantitative Research Group, Arun has worked on stochastic volatility models for derivatives and exotics pricing and hedging (e.g., variance swaps and VIX Futures fair pricing and cross-currency volatility surface construction) and at the intersection of diverse areas such as data science, innovative quantitative finance models across asset classes, and using machine learning methods to help reveal embedded signals in traditional and alternative data that can be used to construct quantitative trading strategies. He holds a PhD in computer science and applied mathematics from Cornell University and a bachelor of technology in computer science from IIT Delhi. Arun lives in central New Jersey with his lovely wife and two children.

Presentations

Trading strategies using alternative data and machine learning 40分钟议题 (40-minute session)

To gain an edge in the markets, quantitative hedge fund managers require automated processing to quickly extract actionable information from unstructured and increasingly nontraditional sources of data. Arun Verma details AI and machine learning (ML) techniques in quantitative finance that lead to profitable trading strategies.

Dingxian Wang is an applied researcher at eBay.

Presentations

基于知识图谱的可解释性推荐系统(Explainable reasoning over knowledge graphs for recommendation) 40分钟议题 (40-minute session)

In recent years, there's been increasing attention on incorporating knowledge graphs into recommender systems. By exploring the interlinks within a knowledge graph, you can discover the connectivity between users and items as paths. Dingxian Wang outlines a new model, knowledge-aware path recurrent network (KPRN), for exploiting knowledge graphs for recommendation.

Long Wang is the vice president of Tencent Cloud and is in charge of the R&D of AI and big data products and services. Previously, he worked in China, Germany, and the US for more than 18 years, serving mainly multinational corporations (MNCs) such as eBay; Siemens; VMware, where he was responsible for VMware’s flagship cloud management product vRealize Automation; and Cheetah Mobile, where he was responsible for the content management system. He has also been founder or cofounder of several startups. He earned his bachelor’s degree from Tsinghua University.

Presentations

云服务加速人工智能创新(Accelerate innovations with AI in the cloud) 主题演讲 (Keynote)

We all know that the cloud is the best place to use new technologies. Long Wang examines what's happening for AI in the cloud: How does AI in the cloud accelerate the innovations in the industry? What's mostly possible? What's still on the way? How does the cloud help?

Tiezhen Wang is a senior software engineer at Google.

Tiezhen Wang Google 软件工程师 TensorFlow 中国团队成员。

Presentations

令人兴奋的TensorFlow 2.0新功能(Exciting new features in TensorFlow 2.0) 40分钟议题 (40-minute session)

TensorFlow 2.0 is a major milestone with a focus on ease of use. Tiezhen Wang walks you through the new exciting features and best practices. Join in to explore distributed strategies and edge deployment (TensorFlow Lite and TensorFlow.js).

Yang Wang is a machine learning engineer on the data analytics team at Intel, focusing on deep learning infrastructure, algorithms, and applications. He’s one of the core contributors of Analytics Zoo and BigDL.

Yang Wang是英特尔数据分析团队的机器学习工程师,专注于深度学习基础架构、算法和应用。他是Analytics Zoo和BigDL的核心贡献者之一。

Presentations

Analytics Zoo:基于Apache Spark的分布式TensorFlow和Keras(Analytics Zoo: Distributed TensorFlow and Keras on Apache Spark) 3小时辅导课 (3-hour Tutorial)

Zhichao Li, Kai Huang, and Yang Wang show you how to build and productionize deep learning applications for big data using Analytics Zoo—a unified analytics and AI platform that seamlessly unites Spark, TensorFlow, Keras, and BigDL programs into an integrated pipeline—illustrated though real-world use cases from JD.com, MLSListings, the World Bank, Baosight, and Midea/KUKA.

Analytics Zoo:基于Apache Spark的生产级别分布式TensorFlow(Analytics Zoo: Distributed TensorFlow in production on Apache Spark) 40分钟议题 (40-minute session)

Building a model is fun and exciting; putting it to production is a different story. Yang Wang offers an overview of Analytics Zoo, a unified analytics and AI platform for distributed TensorFlow, Keras, and BigDL on Apache Spark, designed for production environments. See how you can benefit from its easy deployment, high performance, and efficient model serving for deep learning applications.

王奕恒是腾讯云的高级研发工程师,主要方向是分布式机器学习,尤其是基于Apache Spark构建大规模数据分析平台。他还是Apache Spark上深度学习框架BigDL的主要贡献者。奕恒之前工作于Intel和摩根士丹利。

Presentations

Sparkling: 基于Apache Spark进行一站式机器学习 40分钟议题 (40-minute session)

机器学习项目在企业中实际落地往往涉及到复杂工作流构建和数据管理,以及多种工具的整合。而且随着数据规模的增加,团队规模的扩大,这一任务更具挑战性。Apache Spark是业界流行的大数据框架,被广泛的应用在海量数据的分析处理。本议题将介绍我们在腾讯云上如何基于Apache Spark为客户建立一个一站式机器学习平台的相关工作。主要内容包括多种数据源的接入,构建复杂数据管线,利用数据可视化理解数据,通过可插拔的机制使用各种流行的机器学习框架,以及部署和监控模型。我们也会分享在这一过程中遇到的问题和挑战。听众也可以了解到,通过这种和大数据紧密结合的一站式机器学习,用户可以怎样更加高效的建立和管理他们的机器学习项目,从而加速了机器学习在业务中的落地。

Pete Warden is the technical lead of the mobile and embedded TensorFlow Group on Google’s Brain team.

Presentations

TensorFlow lite for microcontrollers 40分钟议题 (40-minute session)

Pete Warden explains how to use Google's open source framework to run machine learning models on embedded processors like microcontrollers and DSPs. Discover what you need to get started using the code itself, including hardware, coding tools, and getting the library built.

The future of machine learning is tiny 主题演讲 (Keynote)

There are over 250 billion embedded devices in the world. On-device machine learning gives us the ability to turn wasted data into actionable information and will enable a massive number of new applications over the next few years. Pete Warden digs into why embedded machine learning is so important, how to implement it on existing chips, and some of the new use cases it will unlock.

Abigail Hing Wen is head of emerging tech for the Office of the CTO of AI Products at Intel. At Intel Capital, she partners closely with investors on AI investments and has worked with over 100 Silicon Valley startups from incorporation to acquisition or IPO. She advises board members, and writes and speaks on AI and venture capital at international conferences. (See her articles "Fortune on AI & Privacy" and "Forbes on AI and Bias" and highlights from her talk at Hub.Berlin.) She previously practiced law at Sullivan & Cromwell and worked on Capitol Hill. Abigail holds a BA in government from Harvard, a JD from Columbia, and an MFA in writing. Her debut novel is forthcoming Feb. 4, 2020.

Abigail Hing Wen是英特尔人工智能产品部CTO办公室的新兴技术主管。在英特尔投资公司,她和人工智能投资部的投资者进行密切合作,并与100多家硅谷创业公司(从注册成立的,到被收购的或IPO的)进行合作。她为董事会成员提供建议,并在国际会议上就人工智能和风险投资做演讲和撰写文章。她曾在Sullivan&Cromwell律师事务所和Captiol Hill律师事务所工作。Abigail拥有哈佛大学政府专业学士学位、哥伦比亚大学法学博士学位和写作硕士学位。她的首本小说将于2020年2月4日上市。推特账号:@abigailhingwen

《财富》,关于人工智能和隐私: http://fortune.com/2018/12/27/ai-privacy-innovation-machine-learning/

《福布斯》,关于人工智能和偏见(点击量超过15000): https://www.forbes.com/sites/intelai/2019/03/27/the-biggest-ai-ethical-issues-businesses-need-to-address-nowand-how/

Hub.Berlin演讲的重点总结:https://www.intel.ai/videos/the-present-and-future-of-ai-innovations-investments-and-strategies/#gs.7e9o7b

Presentations

Top AI breakthroughs you need to know about 主题演讲 (Keynote)

Abigail Hing Wen catches you up on some of the most exciting recent breakthroughs in the industry, including natural language processing strong enough to generate sentences indistinguishable from a human’s, highly accurate 3D protein structure prediction to fight disease, and leaps forward in reinforcement learning, a more natural but very complex alternative to other forms of machine learning.

Frank Wu is Dell Technologies Vice President and General Manager of Strategy and Business Development for Greater China. In this role he is responsible for shaping Dell’s China 4.0 Strategy, driving its execution and developing new business models, partnerships and application services.

With 20 years’ experience in the IT & telecom industry, Mr. Wu previously served as VP of Alcatel-Lucent Enterprise and General Manager for the Greater China Region, responsible for regional business operation and investment strategy and alliances. Prior to joining Alcatel-Lucent Enterprise, he served as GM of Tellabs China, in charge of business operations in China. In this position, he led his team in making major inroads in the Internet and carrier market sectors Mr. Wu has also held various leadership positions in Nortel Networks, including GM of MEN solutions for Greater China, Head of Data Network, Head of Consulting and Professional Services, and Manager of Global Product Line and Business Planning, based in Canada.

Presentations

解锁数据的力量; 拥抱智能+(由Dell Technologies赞助)(Unlock the power of data; embrace intelligent+ (sponsored by Dell Technologies)) 主题演讲 (Keynote)

This is the data era. Data helps to make better products and services, allowing a company to attract more customers, which results in more data—and repeat. Eventually, this turns into data capital, the most valuable corporate asset. Frank Wu explains why how you use your data will determine your future.

Mingxi Wu is the vice president of engineering at TigerGraph, a Silicon Valley-based startup building a world-leading real-time graph database. Over his career, Mingxi has focused on database research and data management software. Previously, he worked in Microsoft’s SQL Server Group, Oracle’s Relational Database Optimizer Group, and Turn Inc.‘s Big Data Management Group. Lately, his interest has turned to building an easy-to-use and highly expressive graph query language. He’s won research awards from the most prestigious publication venues in database and data mining, including SIGMOD, KDD, and VLDB, and has authored five US patents with three more international patents pending. Mingxi holds a PhD specializing in both database and data mining from the University of Florida.

Presentations

非监督学习在大规模图谱上的案例应用和开源算法剖析 40分钟议题 (40-minute session)

图数据上的非监督学习在激活大数据的经济价值上有着广泛和不可替代的作用。 PageRank能够发掘重要的实体, 社区发掘(community detection)可以找到具有某种特性的群体,紧密度中心性算法(Closeness Centrality)可以自动找到远离群体的个体。所有这些算法都是非监督的学习。 我们分享一些具体客户案例来展示他们的价值,同时分享怎样在大数据上灵活应用这些开源算法。

夏磊先生, 现任英特尔中国人工智能技术架构师,服务于英特尔数据中心技术销售部。专注于为客户在应用人工智能前沿技术过程中为客户的创新提供技术建议与指导提供,并提供英特尔产品与技术相关的支持。
夏磊先生于2000年加入英特尔,历任网络系统工程师、客户技术经理、渠道技术总监、云计算方案架构师、物联网端到端方案架构师,支持了国内信息产业在在互联网、数据中心、云计算与物联网术时代的持续技术创新。
夏磊先生获有机器人工程学士学位。在加入英特尔前任职于政府与教育行业的不同的技术开发和技术教育岗位,在软件算法、自动控制及工程管理等领域具有丰富经验。

Presentations

Intel架构的低精度推断(Low-precision inference on Intel architecture) 40分钟议题 (40-minute session)

Vector neural network instructions (VNNI) is the new Intel instruction set for low-precision AI inference inside the next-generation Xeon platform. Lei Xia offers an overview of the VNNI and Intel software tools, helping you use this new instruction set to accelerate inference with INT8.

Vincent Xie (谢巍盛) is the Chief Data Scientist/Senior Director at Orange Financial, as head of the AI Lab, he built the Big Data & Artificial Intelligence team from scratch, successfully established the big data and AI infrastructure and landed tons of businesses on top, a thorough data-driven transformation strategy successfully boosts the company’s total revenue by many times. Previously, he worked at Intel for about 8 years, mainly on machine learning- and big data-related open source technologies and productions.

Presentations

中国电信如何利用对抗性自动编码器来对抗金融诈骗(How China Telecom combats financial fraud with adversarial autoencoders) 40分钟议题 (40-minute session)

Weisheng Xie dives deep into how China Telecom uses adversarial autoencoders (AAEs) for risk factors modeling to fight a special kind of financial fraud. It's just one step in a long path of unsupervised tasks, but it's proved to be efficient and effective in practice.

Hui Xue is an associate researcher in the System Group at Microsoft Research Asia (MSRA). Her interests include automated machine learning (AutoML), deep learning, and natural language processing, especially the applications for chatbots. She holds a master’s degree in natural language processing from Peking University.

Presentations

自动机器学习(Automated machine learning)技术的实践与应用 40分钟议题 (40-minute session)

人工智能在过去的几年里飞速发展,但是机器学习的实践和应用需要消耗一定的人力和时间。例如,如何去做特征选择,如何设计一个适合该任务的神经网络模型等等。而自动机器学习技术,可以帮助开发者和机器学习实战者,缩短开发周期,提高效率。我们的介绍主要包括:自动机器学习技术的进展;我们开源的自动机器学习开源库neural network intelligence; 如何利用自动机器学习的技术,在产品和应用上提高效率,节省所需的时间和缩短周期。我们会在最后一部分,分享一些利用自动特征选择,自动参数调整以及模型架构搜索上的成功案例。

Season Yang is an analytics fellow in McKinsey & Company’s risk practice. Previously, Season was a data scientist in residence at the Data Incubator, where he also contributes to curriculum development and instruction, and worked at NASA’s Goddard Space Flight Center, where he studied climate change models with data analysis. Season holds a double bachelor’s degree in applied mathematics and scientific computation and economics from UC Davis and a master’s in applied mathematics from Columbia, specializing in numerical computation.

Presentations

Deep learning with TensorFlow 2天培训 (2-day Training)

The TensorFlow library contains computational graphs with automatic parallelization across resources, which is ideal architecture for implementing neural networks. Season Yang introduces TensorFlow's capabilities in Python, and you'll then get your hands dirty building machine learning algorithms piece by piece while using the Keras API provided by TensorFlow with several hands-on applications.

Shan Yu is a machine learning software engineer at Intel.

Intel机器学习软件工程师。

Presentations

基于Spark使用AI来玩游戏(Game playing using AI on Spark) 40分钟议题 (40-minute session)

Using AI to play games is often perceived as an early step toward achieving general machine intelligence, as the ability to reason and make decisions based on sensed information is an essential part of general intelligence. Shan Yu shares lessons learned from her attempts using AI on Spark to play games.

袁理 深圳普思英察科技有限公司 项目及产品总监

袁理拥有AI行业及金融IT行业工作10多年经验,2006年加入汇丰银行环球技术中心。2013年袁理作为汇丰银行风控部门对公信贷风险业务的资深技术架构师及IT项目经理,主要带领印度,中国及香港团队及协调美国、英国、法国团队支持汇丰银行核心及风控等系统研发升级、自动化和敏捷转型以及云端移植可行性探索,2017年袁理加入普思英察至今主要负责AI及无人车行业产品及项目落地以及解决方案预研及商业模式设定等主要工作。

Presentations

自动驾驶技术是如何应用于新潮传媒、新零售行业 40分钟议题 (40-minute session)

如何令自动驾驶技术落地并结合新潮传媒以及新零售业务,相关的技术是如何实现,商业模式是什么以及如何通过人工只能技术提升行业的效率。

Henry Zeng is a principal program manager on the AI platform team at Microsoft, where he works with the engineering team, partners, and customers to ensure AzureML is the best ML platform in the cloud. He’s been in the AI and data area for more than 14 years in areas such as database, big data, machine learning, and deep learning. Previously, he was the lead AI solution architect at Microsoft China, where he worked with partners and customers to land AI solutions in manufactory, retail, finance, education, and public service. Henry holds an MS in computer science from Wuhan University.

Presentations

ONNX:开放和互操作平台让AI无处不在(AI everywhere: Open and interoperable platform for AI with ONNX) 40分钟议题 (40-minute session)

An open and interoperable ecosystem enables you to choose the framework that's right for you, train at scale, and deploy to cloud and edge. ONNX provides a common format supported by many popular frameworks and hardware accelerators. Henry Zeng, Klein Hu, and Emma Ning introduce you to ONNX and its core concepts. (Presented in English and Chinese.)

基于深度学习的时间序列预测 (Deep learning for time series forecasting) 3小时辅导课 (3-hour Tutorial)

Almost every business today uses forecasting to make better decisions and allocate its resources more effectively. Deep learning has achieved a lot of success in computer vision, text, and speech processing but has only recently been applied to time series forecasting. Join in to learn how and when to apply deep neural networks to time series forecasting. (presented in Chinese and English)

通过自动化机器学习民主化和加速AI落地 (Democratizing and accelerating AI through automated machine learning) 3小时辅导课 (3-hour Tutorial)

Intelligent experiences powered by AI seem like magic, but developing them is cumbersome, involving a series of time consuming sequential and interconnected decisions along the way. What if you had an automated service that could identify the best machine learning pipelines for your given problem or data? Lu Zhang, Henry Zeng, and Xiao Zhang detail how automated machine learning does that.

Alina Zhang is a data scientist at Skylinerunners, where she drives the company to provide AI powered Chatbot for online stores, for example, online restaurants booking, grocery shopping, hotel booking, customer service, FAQ.
Previously, Alina was a data scientist helping startups (Nobul, superQuery) to apply machine learning models on user behaviour analysis, recommendation systems, sentiment analysis, classification, and time series forecasting. She worked as a software developer and WLM component owner of IBM DB2 at IBM. She’s a certified Google Cloud Professional Data Engineer and the author of machine learning articles on medium.com/@alina.li.zhang.
Alina holds a masters degree in computer science from Western University, where her research focused on high-performance computing and truncated Fourier transform.

Presentations

Using deep learning and time series forecasting to reduce transit delays 40分钟议题 (40-minute session)

Toronto is unique among North American cities for having a legacy streetcar network as an integral part of its transit system. This means streetcar delays are a major contributor to gridlock in the city. Learn about applying deep learning time series forecasting to machine learning as Mark Ryan and Alina Li Zhang explain how streetcar delays can be predicted...and prevented.

张晨,浙江创邻科技有限公司 创始人,全国五个获评2017年度“中国留学生回国创业启动支持计划”重点类项目人才之一,杭州市全球引才“521”计划专家,浙江省钱江人才计划、杭州市西湖区325海外引才计划A类项目人才,加拿大滑铁卢大学计算机科学博士、麦吉尔大学计算机科学博士后。曾担任美国运通大数据科学家,参与研发美国运通内部第一批大数据挖掘与机器学习的风控算法;其博士论文中基于 HBase 的分布式事务处理的论文和谷歌基于 Bigtable 的分布式事务处理科研同时独立发表,成为了硅谷初创 Splice Machine 的核心底层技术之一,因此受邀成为软件架构师,研发了世界首款基于 Hadoop 的关系型数据库 Splice Machine; 加拿大 Graph Intelligence 联合创始人,搭建了世界首款深度集成于 Hadoop 的原生分布式图数据库。回国后创立浙江创邻科技有限公司,研发了自主知识产权的分布式实时图数据库Galaxybase和创邻天机认知计算平台。获评浙商青云榜30强。

Presentations

创邻Galaxybase图数据库和AI应用(由创邻科技赞助)(The world's fastest graph database Galaxybase and AI applications (sponsored by Chuang Lin Tech)) 40分钟议题 (40-minute session)

浙江创邻科技有限公司创始人张晨将介绍创邻科技自主知识产权的核心技术分布式图数据库Galaxybase。Galaxybase是目前世界上最快、延展性最好的图数据库,比Neo4j快20-100倍,高并发实时读写快1000倍,填补了我国图数据存储及处理领域的空白,并打造了国内首个专注图挖掘的认知计算平台。核心团队由海归大数据专家、国家青年千人、浙江省千人专家、杭州市特聘专家,及国内外名校博士、硕士组成,在海量数据并发并行处理、人工智能、图运算等领域有多项世界领先的技术储备。2018年获得了百度风投BV投资。演讲将介绍图数据库的技术背景、经典应用、Galaxybase的技术、应用和未来的挑战。

Le Zhang is a data scientist with Microsoft Cloud and Artificial Intelligence, where he applies cutting-edge machine learning and artificial intelligence technology to accelerate digital transformation for enterprises and startups on cloud. He’s helped numerous corporations develop and build enterprise-grade scalable advanced data analytical systems with a broad spectrum of application scenarios like manufacturing, predictive maintenance, financial services, ecommerce, and human resource analytics. Le specializes in cloud computing, big data technologies, and artificial intelligence. He enjoys sharing knowledge and learning from people and is a frequent speaker at industrial and academic conferences and community meetups. He holds a PhD in computer engineering.

Presentations

Best practices for building enterprise-grade recommendation systems on Azure with Microsoft/Recommenders 40分钟议题 (40-minute session)

Enterprises benefit from recommendation systems for revenue and customer engagement, but creating such a system is time-consuming. Le Zhang and Jianxun Lian explore the Microsoft/Recommenders repository, which offers solutions to building recommendation systems. It contains classic and state-of-the-art algorithms from Microsoft and enables enterprise success by leveraging Azure's cloud capability.

Lu Zhang is a senior program manager from the AI platform team in the Microsoft Cloud and AI Group. Lu has been working in big data and AI for the past seven years as the product manager of the trustworthy data and data catalog services empowering tens of thousands users in Microsoft to make data-driven decisions. Her current focus is to build the best-in-class ML platform in Azure. Lu holds an MS in computer science from Tsinghua University.

Presentations

通过自动化机器学习民主化和加速AI落地 (Democratizing and accelerating AI through automated machine learning) 3小时辅导课 (3-hour Tutorial)

Intelligent experiences powered by AI seem like magic, but developing them is cumbersome, involving a series of time consuming sequential and interconnected decisions along the way. What if you had an automated service that could identify the best machine learning pipelines for your given problem or data? Lu Zhang, Henry Zeng, and Xiao Zhang detail how automated machine learning does that.

Maria is vice president of engineering at LinkedIn Talent Solutions (LTS) and Careers, which helps recruiters connect with quality talent and connects job-seekers with opportunity. Previously, Maria was CTO at Tinder, where she built a world-class team of engineers and scaled the app to serve a rapidly growing global user base; was vice president of engineering for Yahoo Mobile; managed teams at Microsoft, Zillow, and NetIQ; and founded Alike, a mobile local recommendation app (acquired by Yahoo). She studied computer science at Tsinghua University and holds bachelor’s and master’s degrees from Eastern Michigan University.

Presentations

The future of hiring and the talent market with AI 主题演讲 (Keynote)

If the most dramatic headlines were true, we’d all be preparing for robots to take over our jobs, our lives, and, eventually, the world. But the truth is, automation and AI are doing more to improve the quality of our work than they are to replace us. Maria Zhang examines AI and its impact on people’s jobs, quality of work, and overall business outcomes.

Harper Zhang is a program manager from the Microsoft AI platform team. She works with customers, engineers, and partners to make Azure Machine Learning the best ML choice in the cloud. She holds an MS in communication engineering from Beijing University of Posts and Communications.

Presentations

通过自动化机器学习民主化和加速AI落地 (Democratizing and accelerating AI through automated machine learning) 3小时辅导课 (3-hour Tutorial)

Intelligent experiences powered by AI seem like magic, but developing them is cumbersome, involving a series of time consuming sequential and interconnected decisions along the way. What if you had an automated service that could identify the best machine learning pipelines for your given problem or data? Lu Zhang, Henry Zeng, and Xiao Zhang detail how automated machine learning does that.

Zhen Zhao is a technical consulting engineer at Intel, providing technical consulting and training for AI software solutions, including Intel MKL/MKL-DNN, Intel OpenVINO, and Intel performance libraries (IPP/MKL/DAAL) to Intel strategic customers in the Asia-Pacific, enabling Intel internal and external customers to be successful with Intel platform use through Intel software technology and products.

Zhao为英特尔亚太区战略客户提供英特尔人工智能软件解决方案的技术咨询和培训,包括英特尔MKL/ MKL-DNN、英特尔OpenVINO和英特尔性能库(IPP / MKL / DAAL)。他通过使用英特尔的软件技术和产品,帮助内部和外部客户通过英特尔的平台取得成功。

Presentations

Intel OpenVINO:加速从边缘到云端的深度学习的推断和计算机视觉(Intel OpenVINO: Accelerating deep learning inference and computer vision from edge to cloud) 3小时辅导课 (3-hour Tutorial)

Intel OpenVINO provides a highly optimized cross-platform deep learning deployment and visual AI solution based on various Intel architectures. Join Zhen Zhao as she explains the structure and workflow of the Intel OpenVINO toolkit, optimization methods by asynchronies, heterogeneous computing, low-precision inference, and instruction set acceleration.

Hao Zheng is the cofounder and CTO of PlusAI, a leading global self-driving technology company enabling autonomous commercial trucking fleets. Previously, he was the senior director and lead architect of the mobile search and content personalization at Yahoo, head of the Yahoo Research Lab in Beijing, and Asia CTO of Zynga. Hao had a long tenure in the internet industry and founded two successful startup companies in social analytics and mobile advertising. He earned his master of science degree in computer science and electrical engineering from the University of Wisconsin-Madison and attended the PhD program in electrical engineering at Stanford University.

Presentations

自驾驶技术与未来自动化车辆仓到仓运输(Self-driving technology and the future autonomous depot-to-depot transport) 主题演讲 (Keynote)

PlusAI is developing a full stack self-driving technology to enable large-scale autonomous commercial fleets. Hao Zheng examines some of the unique challenges across different layers of the technology stack of building an autonomous truck that's both safe and efficient and dives into how PlusAI is addressing them.

Presentations

基于数据中心基础架构的深度学习(由Dell Technologies赞助)(A deep learning harness built on data center infrastructure (sponsored by Dell Technologies)) 40分钟议题 (40-minute session)

Improve the utilization rate of data center resources. Join in to explore DL infrastructure and a GPU-as-a-service solution. You'll learn how it simplifies the AI compute requirements with automated access, better control, and simplified provisioning all while pushing your GPU resources to the limit accelerating your model training and inference.

Siyuan Zhuang is a PhD student in RISELab at UC Berkeley. He is interested in the intersection of AI and systems.

Presentations

Building reinforcement learning models and AI applications with Ray 3小时辅导课 (3-hour Tutorial)

Ray is a general-purpose framework for programming your cluster. Richard Liaw leads a deep dive into Ray, walking you through its API and system architecture and sharing application examples, including several state-of-the-art AI algorithms.

Weiqiang Zhuang is a senior software engineer in IBM’s Open Source Data and AI Group focusing on building a cloud native pipeline solution for AI workflows. He was the tech lead of the BigR machine learning project built on top of Hadoop and has contributed to Apache Spark, MLflow, Kubeflow, Apache SystemML, and R4ML. He was also one of the core engineers for DB2’s process model component.

Presentations

AI pipelines on container platforms 40分钟议题 (40-minute session)

AI pipelines simplify the lifecycle workflow management and enhance reproducibility and collaboration for machine learning and deep learning projects. Cloud native platform solutions offer great portability and scalability. Weiqiang Zhuang and Huaxin Gao show how, by combining strengths, AI pipelines on container platforms can help accelerate AI application development and deployment.

刘怀军

美团研究员,美团外卖个性化技术负责人,负责外卖个性化搜索、排序和推荐工作。曾为腾讯搭建公司第一个智能反垃圾系统和智能问答系统,并负责搜搜查询分析,微信智能对话系统和微信搜索算法团队。发明专利20多篇,大部分已经授权。任中文信息学会社会媒体处理专委。

Presentations

AI技术在外卖个性化场景中的落地与思考 40分钟议题 (40-minute session)

该议题的内容包括: 1.外卖个性化场景:个性化搜索,个性化推荐 2.个性化产品形态包括:商家、商品、套餐等 3.外卖个性化中应用的AI技术包括:NLP,DNN,图像技术,强化学习 4.针对外卖业务的特点,介绍个性化场景中,几项重点AI技术的落地、挑战与思考

刘祁跃,爱奇艺智能平台部视频分析负责人,负责视频分析相关算法,包括短视频标签、行为识别、场景识别、目标检测、台词分析、音频分类等,以及视频精彩度分析和智能创作

Presentations

视频精彩度分析及智能创作 40分钟议题 (40-minute session)

对视频进行精彩度分析,有助于筛选优质内容,尤其是冷启动阶段 同时,基于算法对精彩内容的理解,可以辅助创作,如进行标题辅助生成、动态/精彩封面生成、智能拆条等 我们通过对视频、音频、文本等多模态内容分析,同时利用用户交互数据,建立了完备的视频精彩度分析系统,并落地在长/短视频的不同业务场景下,明显提升了业务产出质量和效率

姜涛,音乐检索(MIR)技术专家,有多年从业经验。

Presentations

AI“美颜”你的歌声和视频:K歌修音和自动作曲 40分钟议题 (40-minute session)

介绍如何综合应用多项人工智能技术进行K歌修音和短视频自动配乐,涉及的相关技术包括:人声/音乐分离、高精度的基频提取、自动作曲/作词技术、基于视频内容的音乐生成等。

中国地质大学北京与中国地质科学院联培在读研究生,研究课题是深度学习在地质学上的相关应用!

Presentations

基于目标检测的智能化成矿异常信息提取 40分钟议题 (40-minute session)

矿床所在的位置往往伴随着地质、地球物理、地球化学、遥感异常,因此,这些异常所在的位置也往往伴随着矿床的存在。所以,在找矿工作当中,一个重要的过程便是在地、物、化、遥数据中寻找异常,并将其整合,得出该区域成矿的概率,从而推断出靶区所在的位置。但传统方法并未考虑空间中点与点之间的相关关系。而卷积神经网络中的卷积和池化方法,充分考虑了点与点之间的相关关系。但单纯使用卷积神经网络只能进行特征提取,不能圈定异常所在的区域。因此,特将目标检测的相关算法引入其中,从而圈定异常所在的区域。

目前在阿里巴巴计算平台事业部PAI团队负责大规模深度学习算法基础设施相关建设工作,对大规模分布式机器学习的开发、建设、优化以及在不同业务场景中的落地应用有较为深入的理解和认识。之前先后在奇虎360担当广告技术部门架构师,Yahoo北京研发中心担当效果广告系统技术负责人。

Presentations

PAI张量加速器和优化器:又一个深度学习编译器(PAI tensor accelerator and optimizer: Yet another deep learning compiler) 40分钟议题 (40-minute session)

本次演讲会介绍阿里计算平台PAI团队过去一年多时间里在深度学习编译器领域的技术工作进展----PAI TAO(Tensor Accelerator and Optimizer)。PAI-TAO采用通用编译优化技术,来解决PAI平台所承载的多样性AI workload面临的训练及推理需求的性能优化问题,在部分workload上获得了20%到4X不等的显著加速效果,并且基本作到用户层全透明,在显著提升平台效率性能的同时也有效照顾了用户的使用惯性。目前PAI-TAO已经先后用于支持阿里内部搜索、推荐、图像、文本等多个业务场景的日常训练及推理需求。

杨博理,现任宜信大数据创新中心首席量化科学家,负责宜信线上财富管理平台上的量化投资策略研发、财务规划系统构建、以及AI在财富管理应用层面上的探索。华中科技大学博士后、博士,剑桥大学联合培养博士,里昂高等商学院访问学者。《量化炼金术——中低频量化交易策略研发》一书的作者。

Presentations

线上财富管理领域中的AI应用 40分钟议题 (40-minute session)

AI技术是线上财富管理领域中不可或缺的一环。在这个演讲中,我会将财富管理进一步细分为投资和实现财务目标两个方面,并分别讲解AI技术在这两个细分层面上的应用问题。对于投资而言,一些具备强金融逻辑的变量可能更适合使用机器学习进行预测。而在资产价格的预测上,可以尝试使用AI和大数据技术获取更多的有价值信息。对于实现财务目标而言,基于NLP技术的语义理解、引导式对话是理解用户的关键,基于AI和大数据的KYC也是判断用户状态的有效工具,而一个融合了财务规划、投资和精算知识的专家系统则是定制级规划的核心。

温浩,云从科技联合创始人。2003年获得中国科大电子科学与技术专业学士,并保送中国科大中科院量子信息重点实验室硕博连读,师从“量子调控”973首席科学家郭光灿院士,专攻量子通信器件和网络方向。2008年获得中国科大通信与信息系统博士学位,2014年加入中国科学院重庆绿色智能技术研究院。2015年和周曦博士共同创立云从科技。

Hao Wen is a cofounder, along with Xi Zhou, of CloudWalk Technology. Previously, he worked for the Chongqing Institute of Green Intelligent Technology of the Chinese Academy of Sciences. He holds a bachelor’s degree in electronic science and technology from China University of Science and Technology, where he was recommended for admission into the doctoral program at the China University of Science and Technology’s Key Laboratory of Quantum Information; he completed his PhD in communication and information systems under Guangcan Guo, the “Quantum Control” 973 chief scientist. His research interests include quantum communication devices and networking.

Presentations

打造A.I.闭环 引领产业变革 40分钟议题 (40-minute session)

AI企业发展应该是一个从学术研究、行业验证、商业落地、行业平台到智能生态的一层层深入过程,这也是人工智能企业理想的发展阶段。 云从科技计划打造核心技术闭环,让计算机更好地服务人类。并将全面降低人工智能准入门槛,让“AI普惠”成为可能。

王书浩是透彻影像的联合创始人、技术总监,博士毕业于清华大学,清华大学交叉信息研究院博士后、助理研究员,曾于百度、NovuMind(异构智能)、京东从事人工智能研究,于EuroSys、ECML等会议发表多篇学术论文。

王书浩有着多年的人工智能实践经历,对深度学习有深入的研究,同时对深度学习在大规模集群的实施具有丰富的经验。

Presentations

人工智能病理影像辅助诊断系统——从方法到落地 40分钟议题 (40-minute session)

病理学是医学诊断的“金标准”,病理报告对于临床医生提供进一步治疗策略至关重要。一位能够独立发病理报告的病理医师需要10年以上的培养周期,我国目前共有约1万名注册在案的病理医师,根据WHO的要求,人才缺口为4-9万人。使用人工智能来辅助病理医师对样本进行诊断,不仅能够大幅提高医师的诊断效率,而且可以减少漏诊,提高诊断准确率。数字化的病理影像能够观察到组织的细胞形态,在最高倍数字扫描时,文件尺寸达到GB量级,需要从人工智能和系统工程的层面去应对这些挑战。在这个演讲中,我们将从人工智能系统的构建方法入手,介绍透彻影像与中国人民解放军总医院在消化道病理影像辅助系统研发过程中的技术细节。同时,我们将分享诊断系统从部署到落地使用的一些经验。

Yurong Chen is a principle research scientist and senior research director at Intel and the director of the Cognitive Computing Lab at Intel Labs China, where he’s responsible for driving cutting-edge visual cognition and machine learning research for Intel smart computing. He’s also the co-owner of Intel Labs visual understanding and synthesis program, driving research innovation in smart visual data processing technologies on Intel platforms across Intel Labs. He led the research and development of deep learning-based visual understanding (VU) and leading face analysis technologies to impact Intel architectures/platforms and delivered core technologies to help differentiate Intel products including Intel RealSense SDK, CV SDK, IOT video E2E analytics solutions, and client apps. He led the team to win the Intel China Award (top team award of Intel China) 2016, Intel Labs Academic Awards (top award of Intel labs), and the Gordy Award 2016, 2015, and 2014 for outstanding research achievements on DL-based VU, multimodal emotion recognition and advanced visual analytics. He’s published over 50 technical papers and holds 10+ issued/pending US/PCT patents and 30+ patent applications. He holds a PhD from Tsinghua University, which he followed with postdoctoral research at the Institute of Software, CAS.

Yurong Chen博士是英特尔公司的首席研究科学家和高级研究总监,以及英特尔中国研究院认知计算实验室主任。目前,他负责推进英特尔智能计算的前沿视觉认知(视觉分析和理解)和机器学习研究。他还是英特尔研究院“视觉理解与合成”项目的共同负责人,主导和推动基于英特尔平台的智能视觉数据处理的技术创新。他领导和推动了基于深度学习的视觉理解以及领先人脸分析技术的研究和开发,以此影响英特尔架构/平台设计,并为英特尔实感技术,计算机视觉软件开发包,移动终端应用和物联网端对端视频分析解决方案提供关键技术(人脸检测识别,物体检测,表情识别,深度模型压缩等)。由于在先进视觉分析、多模态情感识别、及基于深度学习的视觉理解方面取得卓越研究成就,他的团队获得了2016年英特尔中国区最高团队奖——英特尔中国荣誉奖,并连续获得了英特尔研究院2014、2015和2016年度全球最高学术奖-戈登•摩尔奖。Chen博士在中国科学院软件研究所完成博士后研究后,于2004年加入英特尔。他于2002年获得清华大学博士学位。他至今已发表顶级学术论文50余篇,拥有10余项美国/国际专利及30多项专利申请。

Presentations

在边缘实现深度学习 40分钟议题 (40-minute session)

深度学习在许多领域尤其是视觉识别/理解方面取得了巨大突破,但它在训练和部署方面都存在一些挑战。本讲座将介绍我们通过高效CNN算法设计、领先DNN模型压缩技术和创新部署时DNN网络结构优化来解决深度学习部署挑战的前沿研究成果。

陈薇博士,现任排列科技首席科学家,江西互联网金融协会特聘风控专家,博金贷金融科技研究院院长。
之前,陈薇曾任职于Lendingclub (NYSE:LC) 任首席数据科学家,负责风险管理相关技术创新,开创性将机器学习与文本数据挖掘系统引入P2P贷款风险分析,取得非常良好的效果,并极大缩短了研发周期,主导的非传统风险模型与决策算法的研究与开发,使公司风控水准远高于美国传统银行。再之前,陈薇曾任Paypal(NYSE:PYPL)主任信贷分析师,专注线上交易风险识别和分析,尤其是银行交易的风险分析和建模设计,创新性将大数据,人工智能和机器学习运用于风险识别和决策。持有内布拉斯加大学计算机科学系博士学位,清华大学计算机工程系硕士及中国人工智能重点实验室成员,曾担任数个学术期刊评审,发表专业论文数十篇。

Presentations

量化互联网金融信用与反欺诈风控 2天培训 (2-day Training)

您想了解金融企业是怎样利用大数据和人工智能技术来画像个人行为并检测欺诈用户的吗?互联网金融幕后的量化分析流程是怎么杨的?个人信用是怎样通过大数据被量化的?在实践过程中,机器学习算法的应用存在着哪些需要关注的方面?怎样通过图谱分析来融合多维数据,为我们区分正常用户和欺诈用户? 这套辅导课基于清华大学交叉信息研究院开设的一门"量化金融信用与风控分析”研究生课。其中会用LendingClub的真实借贷数据做为案例,解说一些具体模型的实现。

黄铃,慧安金科(北京)科技有限公司创始人、CEO,清华大学交叉信息研究院兼职教授。主要技术背景是人工智能、信息安全和金融风控。他是全球为数不多的同时精通人工智能和计算机安全的顶级专家,在美国加州大学伯克利分校获得计算机科学博士 (2002-2007),师从 Anthony Joseph 和 Michael Jordan ,从事机器学习算法研究以及计算机网络建模应用。他是美国硅谷著名的反欺诈公司DataVisor的创始成员和大数据总监 (2014-1016),主持了公司整个机器学习,用户行为分析和信用分析系统。他在美国英特尔研究院任资深科学家七年(2007-2014),和 Intel McAfee 开展多个合作项目,应用人工智能技术解决网络和数据安全问题。他在人工智能,大数据分析和金融科技相关领域有近十五年的研究和开发背景,在世界顶尖会议上发表近50篇论文,在 Google Scholar 上总引用已超过5,000次。

Presentations

量化互联网金融信用与反欺诈风控 2天培训 (2-day Training)

您想了解金融企业是怎样利用大数据和人工智能技术来画像个人行为并检测欺诈用户的吗?互联网金融幕后的量化分析流程是怎么杨的?个人信用是怎样通过大数据被量化的?在实践过程中,机器学习算法的应用存在着哪些需要关注的方面?怎样通过图谱分析来融合多维数据,为我们区分正常用户和欺诈用户? 这套辅导课基于清华大学交叉信息研究院开设的一门"量化金融信用与风控分析”研究生课。其中会用LendingClub的真实借贷数据做为案例,解说一些具体模型的实现。

目前在阿里巴巴PAI团队负责GPU底层核心优化工作,之前在中科院软件所从事计算机系统结构相关研究工作,对高性能计算、微处理器设计、异构计算领域有较深入的理解和认识,先后有多篇论文在PPoPP、Micro、ACL等体系结构及AI领域顶级会议发表。

Presentations

PAI张量加速器和优化器:又一个深度学习编译器(PAI tensor accelerator and optimizer: Yet another deep learning compiler) 40分钟议题 (40-minute session)

本次演讲会介绍阿里计算平台PAI团队过去一年多时间里在深度学习编译器领域的技术工作进展----PAI TAO(Tensor Accelerator and Optimizer)。PAI-TAO采用通用编译优化技术,来解决PAI平台所承载的多样性AI workload面临的训练及推理需求的性能优化问题,在部分workload上获得了20%到4X不等的显著加速效果,并且基本作到用户层全透明,在显著提升平台效率性能的同时也有效照顾了用户的使用惯性。目前PAI-TAO已经先后用于支持阿里内部搜索、推荐、图像、文本等多个业务场景的日常训练及推理需求。