[2018-04-13] Ed H. Chi, Google,” Recent Advances in Neural Recurrent Recommendations”


Title: Recent Advances in Neural Recurrent Recommendations

Date: 2018-04-13 15:40-17:00

Location: R103, CSIE

Speaker: Ed H. Chi, Google

Hosted by: Prof.Winston Hsu


It has been much harder to obtain the same magnitude of improvement in recommenders compared to recent advances in many long-standing AI problems such as recognition and classifications of images, audio, and videos, as well as machine translation of human languages. Some reasons include: (1) large amount of data making training difficult, yet having (2) noisy and sparse labels; (3) changing dynamics of context such as user preferences and items; and (4) latency requirement for a recommendation response.


In this talk, I will summarize several recent advances in utilizing recurrent neural networks for recommendations.  First, recurrent models are better able to generate recommendation, but improvements here required careful modeling of time, and other contextual features. Second, attentional mechanisms enable the recurrent model to attend to modulate the input and output states. Third, reinforcement learning over large recommendation spaces is possible using policy gradients and off-policy correction. I will conclude by offering some thoughts on deep learning research from looking through the lens of the recommendation problem and its complexity.



Ed H. Chi is a Principal Scientist at Google, leading a machine learning research team focused on recommendation systems and social computing research. He has launched significant improvements of recommenders for YouTube, Google Play Store and Google+. With 39 patents and over 110 research articles, he is known for research on Web and online social systems, and the effects of social signals on user behavior.


Prior to Google, he was the Area Manager and a Principal Scientist at Palo Alto Research Center's Augmented Social Cognition Group, where he led the team in understanding how social systems help groups of people to remember, think and reason. Ed completed his three degrees (B.S., M.S., and Ph.D.) in 6.5 years from University of Minnesota, and has been doing research on software systems since 1993. Recognized as an ACM Distinguished Scientist and elected into the CHI Academy, he has been featured and quoted in the press, including the Economist, Time Magazine, LA Times, and the Associated Press, and has won awards for both teaching and research. In his spare time, Ed is an avid photographer and snowboarder, and has a blackbelt in Taekwondo.

最後修改時間:2018-04-13 PM 2:30

cron web_use_log