Presented By O’Reilly and Intel AI
Put AI to work
April 10-11, 2018: Training
April 11-13, 2018: Tutorials & Conference
Beijing, CN

Schedule: 模型与方法 (Models and Methods) sessions

Add to your personal schedule
09:0012:30 Wednesday, April 11, 2018
Location: 紫金大厅B(Grand Hall B)
Secondary topics:  运输与物流 (Transportation and Logistics)
Erran Li (Uber ATG)
尽管最近人工智能等领域取得了很多的进展,但自动驾驶里的主要问题(不管是基础研究还是工程应用上的挑战)离完全被解决还有很大的距离。Erran Li将会探索自动驾驶所用的机器学习的基础,并讨论目前相关工作的进展。 Read more.
Add to your personal schedule
09:0012:30 Wednesday, April 11, 2018
Location: 报告厅(Auditorium)
Secondary topics:  深度学习(Deep Learning)
Yufeng Guo (Google)
Yufeng Guo walks you through training a machine learning system using popular open source library TensorFlow, starting from conceptual overviews and building all the way up to complex classifiers. Along the way, you'll gain insight into deep learning and how it can apply to complex problems in science and industry. Read more.
Add to your personal schedule
09:0017:00 Wednesday, April 11, 2018
Location: 多功能厅2(Function Room 2)
Secondary topics:  深度学习(Deep Learning)
Zhichao Li (Intel)
深度学习技术的进步继续推动数据分析和机器学习的演变,推进人工智能的全新应用。作为最受欢迎的上层神经网络 API 之一,Keras 可帮助企业轻松、快速地进行原型构建,并支持多个后端,包括 TensorFlow 和 Theano。在本演讲中,我们将展示如何将 Keras 无缝集成在 BigDL(Apache Spark 的一种分布式深度学习框架)中,以便用户通过运行分布式 Keras 在基于英特尔® 至强® 处理器的现有 Hadoop/Spark 集群上进行训练、微调或大规模推理。 Read more.
Add to your personal schedule
13:3017:00 Wednesday, April 11, 2018
Location: 报告厅(Auditorium) Level: Intermediate
Secondary topics:  增强学习(Reinforcement Learning)
Arthur Juliani (Unity Technologies), Leon Chen (Unity Technologies)
Recently, computers have been able to learn to play Atari games, Go, and first-person shooters at a superhuman level. Underlying all these accomplishments is deep reinforcement learning. Arthur Juliani and Leon Chen lead a deep dive into reinforcement learning, from the basics using lookup tables and GridWorld all the way to solving complex 3D tasks with deep neural networks. Read more.
Add to your personal schedule
09:3009:45 Thursday, April 12, 2018
Location: 紫金大厅A(Grand Hall A)
Secondary topics:  计算机视觉(Computer Vision)
Reza Zadeh (Matroid | Stanford)
Reza Zadeh details three challenges on the way to building cutting-edge ML products, with a focus on computer vision, offering examples, recommendations, and lessons learned. Read more.
Add to your personal schedule
11:1511:55 Thursday, April 12, 2018
Location: 多功能厅3A+B (Function Room 3A+B)
Secondary topics:  增强学习(Reinforcement Learning), 运输与物流 (Transportation and Logistics)
Erran Li (Uber ATG)
深度增强学习已经让人工智能体在很多挑战性的领域可以取得超越人类的表现,例如玩Atari的游戏以及下围棋。这一方法还具有能显著地推进自动驾驶的潜力。Erran Li将会讨论近期在模仿学习方面(例如infoGAIL)、策略梯度法和层次增强学习(例如option-critic架构)等方面的进步,以及它们在自动驾驶方面的应用。Erran接着还会介绍在这个领域需要关注的剩余的挑战。 Read more.
Add to your personal schedule
11:1511:55 Thursday, April 12, 2018
Location: 多功能厅5A+B(Function Room 5A+B) Level:
Secondary topics:  深度学习(Deep Learning), 设计AI平台(Designing AI platforms)
杨军 (阿里巴巴)
本议题会分享我们在典型互联网业务场景(图像、文本处理等)下的深度学习优化实践经验,包括离线训练和在线Inference,并会从系统与算法相结合的角度进行相关经验的阐述和介绍。 Read more.
Add to your personal schedule
11:1511:55 Thursday, April 12, 2018
Location: 多功能厅6A+B(Function Room 6A+B)
Secondary topics:  自然语言与语音技术(Natural Language and Speech Technologies)
Yinyin Liu (Intel AI)
自然语言处理(NLP)带给计算机理解人类语言的能力。NLP利用深度学习最新算法发展例如文档理解之类的应用,使公司能够筛查海量文本,分类并找到相关信息。本议题我们将讨论深度学习最新发展如何影响处理文本、语言及基于对话应用,并启发了利用数据的新方向。我们还将讨论几个使用Intel® AI技术的NLP企业案例。 Read more.
Add to your personal schedule
13:1013:50 Thursday, April 12, 2018
Location: 紫金大厅B(Grand Hall B) Level: Intermediate
Secondary topics:  传媒、广告、娱乐 (Media, Advertising, Entertainment), 自然语言与语音技术(Natural Language and Speech Technologies), 金融服务 (Financial Services)
Zhefu Shi (University of Missouri)
It is critical to analyze the business impact of worldwide events on the financial market. Zhefu Shi explains how to use AI to analyze the impact of financial news using a financial data pipeline. Zhefu outlines how to extract financial entity information and use it to analyze business impact. All of the components use AI to enhance functionality. Read more.
Add to your personal schedule
13:1013:50 Thursday, April 12, 2018
Location: 报告厅(Auditorium) Level: Intermediate
Secondary topics:  AI应用的硬件、软件栈(Hardware and Software stack for AI applications)
Tags: wl
Ruiwen Zhang (SAS Institute)
Drawing on several real-world cases, Ruiwen Zhang demonstrates how to visualize the structure of a probabilistic model and provide better insights into the model's properties, which can be further used to design and motivate new models. She also explains how to reduce the computational complexity required to perform inference and learning in sophisticated models using graphical models. Read more.
Add to your personal schedule
13:1013:50 Thursday, April 12, 2018
Location: 多功能厅2(Function Room 2) Level: Intermediate
Secondary topics:  自然语言与语音技术(Natural Language and Speech Technologies)
Emmanuel Ameisen (Insight Data Science), Yan Kou (Insight Data Science)
Emmanuel Ameisen and Yan Kou share a guide for moving your company toward deep learning using a collection of NLP best practices gathered from conversations with 75+ teams from Google, Facebook, Amazon, Twitter, Salesforce, Airbnb, Capital One, Bloomberg, and others. Read more.
Add to your personal schedule
13:1013:50 Thursday, April 12, 2018
Location: 多功能厅5A+B(Function Room 5A+B) Level:
Secondary topics:  自然语言与语音技术(Natural Language and Speech Technologies)
Tie-Yan Liu (微软亚洲研究院 (Microsoft Research Asia))
以深度学习为代表的人工智能技术通常需要大量的有标签训练数据,这对于很多应用领域而言并非易事。为了解决这个挑战,我们利用人工智能的对称之美——很多人工智能任务天然就是双向的,比如中到英翻译 vs.英到中翻译,图像分类 vs. 图像生成,语音识别 vs. 语音合成——来为机器学习建立闭环、生成有效的反馈信号,从而在缺乏有标签数据的情况下也能实现高效学习。我们将这种新型的学习方法称之为“对偶学习”。对偶学习已经被成功应用到诸多领域,取得了非同凡响的效果。本报告中,我们将针对对偶学习的数学模型、优化算法、概率解释、实验结果,收敛性分析等进行详细讨论,展示对偶学习的魅力,并对它在人工智能领域的更广泛应用进行展望。对偶学习有关的研究成果已发表在NIPS、ICML、IJCAI、AAAI等人工智能领域最顶尖的国际会议之上。 Read more.
Add to your personal schedule
13:1013:50 Thursday, April 12, 2018
Location: 多功能厅6A+B(Function Room 6A+B)
Secondary topics:  AI应用的硬件、软件栈(Hardware and Software stack for AI applications)
Mingxi Wu (Tigergraph inc), Yu Xu (TigerGraph)
为了让机器像人一样思考,一个成功的人工智能应用程序的关键部分必须由强大的数据管理软件支持。在这次演讲中,我们将讨论人工智能数据管理的需求,并指出图模型的独特优势。我们将深入讨论几个现实生活中部署的,且将它们的成功归因于图模型的人工智能应用程序。 Read more.
Add to your personal schedule
14:0014:40 Thursday, April 12, 2018
Location: 多功能厅3A+B (Function Room 3A+B)
Secondary topics:  深度学习(Deep Learning), 运输与物流 (Transportation and Logistics)
Bichen Wu (UC Berkeley)
深度学习近年来的成功极大地促进了自动驾驶技术的快速发展。但不少问题依然存在:1)深度学习模型需要大量的训练数据 2)即便是深度学习模型也很难达到100%准确率 3) 深度学习模型的计算复杂度太高,超出了车载计算机的处理能力。这个讲座将会关注以上几个问题。 Read more.
Add to your personal schedule
14:0014:40 Thursday, April 12, 2018
Location: 紫金大厅B(Grand Hall B) Level: Intermediate
Secondary topics:  自然语言与语音技术(Natural Language and Speech Technologies)
David Talby (Pacific AI)
Average rating: *****
(5.00, 1 rating)
Natural language processing is a key component in many data science systems that must understand or reason about text. David Talby offers an overview of the NLP library for Apache Spark, which natively extends Spark ML to provide open source, fully distributed, and optimized versions of state-of-the-art NLP algorithms, covering the library's design and sharing working code samples in PySpark. Read more.
Add to your personal schedule
14:0014:40 Thursday, April 12, 2018
Location: 多功能厅2(Function Room 2)
Secondary topics:  传媒、广告、娱乐 (Media, Advertising, Entertainment), 增强学习(Reinforcement Learning)
Danny Lange (Unity Technologies)
Danny Lange demonstrates the role games can play in driving the development of reinforcement learning algorithms. Danny uses the Unity Engine with the ML-Agents toolkit as an example of how dynamic 3D game environments can be utilized for machine learning research. Read more.
Add to your personal schedule
14:5015:30 Thursday, April 12, 2018
Location: 紫金大厅B(Grand Hall B) Level: Intermediate
Secondary topics:  自然语言与语音技术(Natural Language and Speech Technologies)
David Talby (Pacific AI)
Average rating: *****
(5.00, 1 rating)
To achieve high accuracy when reasoning about text, you generally need to understand specific languages, jargon, domain-specific documents, and writing styles. David Talby explains how to train custom word embeddings, named entity recognition, and question-answering models on the NLP library for Apache Spark. Read more.
Add to your personal schedule
14:5015:30 Thursday, April 12, 2018
Location: 多功能厅6A+B(Function Room 6A+B) Level:
Secondary topics:  AI应用的硬件、软件栈(Hardware and Software stack for AI applications), 设计AI平台(Designing AI platforms)
Xiaolei Xu (上海新智新氦数据科技有限公司)
目前单机多卡训练是深度学习的标配,但是单机的GPU数目总有上限,因此如何通过多机多卡进行高效的分布式训练就尤其重要。比如,如何将简单的单机程序快速部署到多机并得到相应的加速比,如何使得对GPU的调度与大数据处理平台无缝对接,并使GPU成为平台上按需调度、动态扩容的资源,这些问题的解决对算法迭代优化起到关键作用。 本次talk会详细介绍如何基于Kubernetes和Docker构建TensorFlow的微服务化应用,具体从以下几个方面展开:从少量样本数据的单机快速原型设计验证,无缝切换到大量全数据的多机多卡分布式训练过程;一键启动分布式训练,即基于新氦定制的深度学习云平台,用户无需关注分布式细节,可直接通过可视化web界面进行分布式参数配置和训练代码提交,并可实时可视化监控模型训练收敛性、系统资源消耗和模型输出日志等;模型训练结束后可实时serving将模型快速部署到生产环境。 Read more.
Add to your personal schedule
16:2017:00 Thursday, April 12, 2018
Location: 多功能厅2(Function Room 2)
Secondary topics:  保健与医疗 (Health and Medicine)
Arjun Bansal (Intel)
Precision medicine promises to revolutionize healthcare by delivering better health outcomes at lower cost by eliminating trial-and-error medicine, and Intel is working to make this a reality. Arjun Bansal shares emerging algorithms and models used to analyze healthcare data, including electronic health records, medical images, and pharmaceutical and genomics datasets. Read more.
Add to your personal schedule
16:2017:00 Thursday, April 12, 2018
Location: 多功能厅6A+B(Function Room 6A+B) Level:
Secondary topics:  AI应用的硬件、软件栈(Hardware and Software stack for AI applications), 深度学习(Deep Learning)
Xiaohui Wang (TalkingData)
目前,深度学习在移动端的应用越来越受到重视,从芯片制造商到手机厂商,一直到应用开发者,都在为在智能手机上运行深度学习模型做出了很多努,开发者一方面很难找到针对移动端优化过的解决特定应用场景的模型,一方面不知道应该如何选择这些框架,TalkingData 推出的 Android Deep Learning Framework 就为了解决这些问题。我们提供了针对移动平台的各种类型的模型,以及它们在主流机型行的实测 Benchmark,另外也提供了利用这些预训练模型和自己的数据集进行再训练的服务器端脚本和自动化工具,最后就是封装了一个上层 DL API,让开发者可以支持各种移动端深度学习框架,并为这些模型的使用提供统计分析服务。 Read more.
Add to your personal schedule
11:1511:55 Friday, April 13, 2018
Location: 多功能厅3A+B (Function Room 3A+B)
Secondary topics:  计算机视觉(Computer Vision)
Baining Guo (微软亚洲研究院 (Microsoft Research Asia))
关于微软亚洲研究院通过人工智能技术进行图像合成的最新研究概述。从把普通照片变成毕加索风格的绘画,到生成莱昂纳多·迪卡普里奥(Leonardo DiCaprio)的新图像,我们展示了深度学习所带来的新的可能性。 Read more.
Add to your personal schedule
11:1511:55 Friday, April 13, 2018
Location: 紫金大厅B(Grand Hall B) Level: Beginner
Roger Chen (Computable Labs)
Blockchain technologies offer new internet primitives for creating open and online data marketplaces. Roger Chen explores how data markets can be constructed and how they offer a shared resource on the internet for AI-based research, discovery, and development. Read more.
Add to your personal schedule
14:0014:40 Friday, April 13, 2018
Location: 多功能厅3A+B (Function Room 3A+B)
Secondary topics:  深度学习(Deep Learning)
Sherry Moore (Google)
TensorFlow可以让你进行高速运算,很多时候是在机器学习的情景下。 Sherry Moore将会介绍TensorFlow的最新进展,包括TensorFlow立刻执行机制和TensorFlow Lite。她还会分享一些最佳实践,并将演示机器学习的一些有用的应用。 Read more.
Add to your personal schedule
14:0014:40 Friday, April 13, 2018
Location: 报告厅(Auditorium) Level:
周明 (微软亚洲研究院 (Microsoft Research Asia))
创作诗歌、音乐是人类独具的能力。然而,随着深度神经网络和大数据的发展,计算机已经逐步具备了创作诗歌和音乐的能力。我们致力于把AI融入到创作过程中,并且帮助普通实现创作梦想。为此,我们长期以来进行了对联、诗词的研究。2005年就开发了中文对联系统(http://duilian.msra.cn).。以后又陆续开发了格律诗写作,猜字谜和出字谜。2016年开发了小冰写诗。目前我们正在探索先进的神经网络和大数据来模仿人类的音乐创作过程。我们采用了融入上下文的编码-解码方法来产生诗歌、歌词和谱曲。取得了富有希望的成果。我们的电脑音乐创作已经在CCTV的机智过人节目播出。获得好评,由电脑写出歌词,然后配上曲谱,然后通过声音合成,唱出歌曲。 Read more.
Add to your personal schedule
14:0014:40 Friday, April 13, 2018
Location: 多功能厅5A+B(Function Room 5A+B)
Secondary topics:  运输与物流 (Transportation and Logistics)
Liyun Li (京东硅谷研发中心X-lab)
尽管人工智能技术已经在诸如计算机视觉和自然语言处理等领域获得了巨大的成功,如何在自动驾驶系统中有效地利用AI的能力仍然是一个很大的挑战。我们将以"Apollo"这一百度的开源无人驾驶平台系统做为基准和样例, 深入讨论并且分享在搭建智能的无人驾驶系统各个方面利用AI技术的实践和经验。通过讲解Apollo无人驾驶系统背后的设计理念以及各个功能模块,我们将分享并展示AI技术在Apollo无人驾驶系统中各方面的应用, 包括环境感知,行为预测,行为决策,以及控制规划等。同时我们将结合Apollo系统中的端到端学习实践,探讨AI技术在未来无人驾驶系统中更好的应用场景。 Read more.
Add to your personal schedule
14:5015:30 Friday, April 13, 2018
Location: 报告厅(Auditorium) Level:
Secondary topics:  深度学习(Deep Learning), 运输与物流 (Transportation and Logistics)
Li Li (ESRI)
制图学是一个历史悠久的学科。古希腊地理学家C.托勒密的《地理学指南》就是一部地图制图学著作。托勒密认为地理学就是“以线画形式描绘地球上所有迄今已知的部分及其附属的东西”。几百年以来,地图学领域都没有重大突破。 深度学习作为一个新的技术已经渗透到了各个行业。带来了各种各种的技术革新。本讲座就是探讨如何用深度学习来给地图换装。然后展示一些用深度学习技术给地图换装的结果。并讨论,深度学习在制图领域的应用。 Read more.
Add to your personal schedule
16:2017:00 Friday, April 13, 2018
Location: 紫金大厅B(Grand Hall B) Level: Beginner
Hendra Suryanto (Rich Data Corporation )
Hendra Suryanto shares a case study from a Canadian financial lender that his company helped transition from manual to automated credit decisioning, using gradient boosting machine and deep learning to build the model. In addition to modeling techniques, Hendra highlights the role feature engineering plays in improving model performance. Read more.
Add to your personal schedule
16:2017:00 Friday, April 13, 2018
Location: 报告厅(Auditorium) Level:
Secondary topics:  深度学习(Deep Learning)
李苍柏 (中国地质科学院矿产资源研究所)
众所周知,现在的深度学习已经在各个行业开始了应用。但是深度学习如何与地质行业相结合,这还是一个新兴的话题,国外目前,已经开始用深度学习来处理实验室地震数据,用以提高地震预测的时间;国内也已经有很多人用卷积神经网络开始对岩石图像数据进行处理,这次议题我做的报告是,在介绍前人工作的基础上,介绍一下自己在地质上的应用! Read more.