Thursday, January 2, 2020
Machine learning, the process by which a computer uses algorithms and data to make predictions or decisions on its own, has tremendous potential for social and public good. Machine learning has transformed various industries and solved challenging problems in healthcare, education, media, and transportation. However, machine learning presents its own set of challenges and questions, especially when it comes to its societal impacts. William Wang, an assistant professor in UC Santa Barbara’s Department of Computer Science, is part of a movement to establish an ethical foundation for machine learning. Wang, who serves as director of UCSB’s Center for Responsible Machine Learning (CRML), examines important societal factors that should be considered when building algorithms such as fairness, transparency, and accountability. Launched in fall 2019, the CRML reflects the university’s commitment to define and build the future of machine learning algorithms and artificial intelligence (AI) systems.
“The center allows UCSB to advance our academic excellence in the areas of AI, machine learning, natural language processing (NLP), and computer vision,” explained Wang. “A unique emphasis in the center is tying cutting-edge research in AI with important societal impacts. We are also interested in understanding ethical issues in AI research, including legal and policy issues, and energy efficiency.”
The new year will bring with it a new channel for Wang and his center to spread their messages, the CRML Distinguished Lecture Series.
“These will be research-oriented academic lectures that dive into technical discussions about the solutions for the center’s key topics,” said Wang, who has received a DARPA Young Faculty Award from the U.S. Department of Defense, three IBM faculty awards, and multiple research awards from Facebook and Adobe for his work on machine learning and NLP. “Our hope is that these lectures will help increase the depth of research in the center and allow UCSB students and faculty to engage with leaders in the AI community.”
The first distinguished lecture will be given at 3:30 PM on Thursday, January 16, by Dr. Raymond Mooney, a professor in the Department of Computer Science at the University of Texas at Austin. An elected fellow of the Association for the Advancement of Artificial Intelligence (AAAI), the Association for Computing Machinery (ACM), and the Association for Computational Linguistics (ACL), Mooney has authored more than 170 published research papers, primarily in the areas of machine learning and NLP. Mooney also received the Classic Paper Award from AAAI, and best paper awards from AAAI, ACL, and Knowledge Discovery and Data Mining (KDD).
“Raymond Mooney has been a very insightful leader in our field, and he is particularly well known for his critical opinions of key methods in the development of NLP,” said Wang. “He has a proven track record of doing amazing research that tackle some of our field’s most difficult problems.”
During his presentation, “Robots that Learn Grounded Language Through Interactive Dialog,” Mooney will discuss methods he has developed to teach an office or home robot to accept natural language commands through natural dialog, rather than from manually labeled data. By engaging users in dialog, or conversation, the system learns a semantic parser to connect words to multi-modal (visual, auditory, and sense of touch) perception, which allows the robot to extract the precise meaning of an utterance. After testing his approach online with simulated robots and with people interacting with robots in his lab, Mooney’s methods successfully produced shorter dialogs over time and more accurately identified objects from natural language descriptions using multi-modal perception.
“He will discuss one of the most challenging aspects in AI: teaching machines to read, see, and understand our world,” said Wang. “How do humans connect images to concepts and knowledge in our mind? This is a fundamental challenge in our field because we need machines to understand our world, our common sense, and our basic knowledge. Only then, is it possible to have responsible machine learning.”
The lecture will take place at 3:30 PM on Thursday, January 16, inside the Mosher Alumni House. The lecture is free and open to the public. A reception will precede the lecture at 3 PM.
The series will continue with a lecture at 2 PM on Friday, February 28, with Dan Roth, a distinguished professor from the University of Pennsylvania’s Computer and Information Science Department. Roth’s research focuses on the computational foundations of intelligent behavior.
For more information about the CRML, visit http://ml.ucsb.edu/.
Center for Responsible Machine Learning Distinguished Lecture Series
Thursday, January 16, 2020 - 3:30 PM - Mosher Alumni House, UCSB
“Robots that Learn Grounded Language Through Interactive Dialog”
Dr. Raymond Moody
Professor, Computer Science
University of Texas, Austin
Reception to begin at 3 PM
Friday, February 28, 2020 – 2 PM – Location, TBA
Dr. Dan Roth
Distinguished Professor, Computer and Information Science
University of Pennsylvania
Reception to begin at 1:30 PM