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

Turning machine learning research into products for industry

This will be presented in English.

Reza Zadeh (Matroid | Stanford)
09:3009:45 Thursday, April 12, 2018
Secondary topics:  计算机视觉(Computer Vision)

本讲话将用英语授课,同时会提供中文同声传译。中文版本摘要会在英文摘要下面给出。

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.

  1. Since ML is all about approximation, it can be difficult to assess when a research result is good enough for the industry.
  2. Building systems that scale ML models in production is a challenge on its own. Although a model may work in the lab, scaling it to millions of users may be impossible without further research.
  3. Building good user interfaces for ML products is crucial. Since ML researchers often don’t have a background in user-focused design, they tend to underestimate the importance of good UX design.

将最前沿的研究成果转化成现实的产品是一个巨大的挑战,特别是对于机器学习而言。我们会讨论三个构建最前沿机器学习产品正面临的挑战。首先,因为机器学习都是关于近似的,所以很难评估一个研究结果对于商业目标是否足够好。第二,构建一个能把机器学习模型扩展到生产环境的系统本身就是一个挑战。一个模型可能在实验环境里工作得很好,但是没有进一步的研究就想把它扩展到上百万用户的系统里是不可能的。第三,为机器学习系统构建好的用户界面也非常重要。因为机器学习研究人员通常没有以用户为中心进行设计的背景知识,他们会倾向于低估好的UX设计的重要性。这里我们会以计算机视觉产品为例,使用案例、建议和经验教训来探讨这些挑战。

Photo of Reza Zadeh

Reza Zadeh

Matroid | Stanford

Reza Bosagh Zadeh is founder and CEO at Matroid and an adjunct professor at Stanford University, where he teaches two PhD-level classes: Distributed Algorithms and Optimization and Discrete Mathematics and Algorithms. His work focuses on machine learning, distributed computing, and discrete applied mathematics. His awards include a KDD best paper award and the Gene Golub Outstanding Thesis Award. Reza has served on the technical advisory boards of Microsoft and Databricks. He is the initial creator of the linear algebra package in Apache Spark. Through Apache Spark, Reza’s work has been incorporated into industrial and academic cluster computing environments. Reza holds a PhD in computational mathematics from Stanford, where he worked under the supervision of Gunnar Carlsson. As part of his research, Reza built the machine learning algorithms behind Twitter’s who-to-follow system, the first product to use machine learning at Twitter.

Comments on this page are now closed.

Comments

HW Yang | PROF
2018-04-12 14:18 CST

1. As the ML is based on probility, what is the big prob when it would be used to predict the real model in business or science? And its reliability?
2. Cons and Pros of ML, and future development