** You must register on Eventbrite for this event. 50 free passes are available for people signing up early. Register here:
Join us to hear the cutting edge research from Facebook AI Research. Topic: Approach Human Heuristics with Deep Learning and Reinforcement Learning
Yuandong Tian is a Research Scientist and Manager in Facebook AI Research, working on deep reinforcement learning and its applications, and theoretical analysis of deep models. He is the lead scientist and engineer for ELF OpenGo and DarkForest Go project. Prior to that, he was a researcher and engineer in Google Self-driving Car team in[masked]. He received Ph.D in Robotics Institute, Carnegie Mellon University in 2013, Bachelor and Master degree of Computer Science in Shanghai Jiao Tong University. He is the recipient of 2013 ICCV Marr Prize Honorable Mentions.
Deep Learning has achieved great success in recent years, in particular when dealing with natural data input (e.g., computer vision, speech recognition and natural language processing). In contrast, how to apply deep neural networks on structured data (e.g., logs, problem specifications, code) to accelerate critical applications that traditionally require decades of human heuristics still remains an open problem to address. In this talk, I will cover our recent works in which neural networks, coupled with reinforcement learning and search methods, are used to learn heuristics of a complicated optimization problem, to achieve better performance than using human experience. The application includes online job scheduling, vehicle routing, architecture search, and code decompiler.