Deep learning has been in the center of AI since its revival a few years ago and has achieved tremendous success in many applications. However, while deep neural networks has achieved unprecedentedly great performance, their predictions of usually do not come with a reliable and well-calibrated confidence score. Wrong but confident predictions place great threads to critical real-life applications, e.g. self-driving car. On the other hand, a good confidence estimation is also very useful in implementing effective active learning.
This talk is a tutorial of confidence estimation for deep neural networks and recent advances, and comparison of different methods with focus on the following:
Jialin Jiao, an experienced technical veteran in HD maps, autonomous driving, machine learning, search, software engineering, specializes in research and development in the fields of Location-Based Service (LBS), HD maps, big data, machine learning, artificial intelligence and search. He had worked in Uber (US headquarter), Microsoft search engine Bing, IBM TJ Watson Research. He is currently working for Pony.ai, a startup building autonomous driving car. He holds a bachelor’s degree in computer science from Sun Yat-sen University, a master’s degree in computer science from Shanghai Jiao Tong University, and a master’s degree in electronic engineering from the University of Michigan – Dearborn.
Mr. Jiao is an IEEE member, as well as a founding member of IEEE Computer Society’s Special Technical Community in Autonomous Driving (https://stc.computer.org/autonomousdriving/leadership/).
Help us make this conference the best it can be for you. Have questions you'd like this speaker to address? Suggestions for issues that deserve extra attention? Feedback that you'd like to share with the speaker and other attendees?
Join the conversation here (requires login)
©2019, O'Reilly Media, Inc. • (800) 889-8969 or (707) 827-7019 • Monday-Friday 7:30am-5pm PT • All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners. • firstname.lastname@example.org