From Machine Learning to Machine Reasoning
Assistant Professor, Rutgers University
Time: October 16, 2019 @ 11:15 AM to 12:05 PM
Location: 204 Evans Hall
Intelligent decision support systems have been widely embedded into the daily life of human-beings, such as recommendation systems in social networks and e-commerce, search engines that people use nearly everyday, and the emerging conversational agents such as Siri, Alexa, and Google Assistant. They help users to make informed decisions by providing helpful information or insightful suggestions. In recent years, intelligent decision
support systems extensively rely on machine learning (mostly deep learning) methods to train a black-box model over large-scale data for decision making, which — though with high precision in many cases — are less effective in telling us why a particular decision is made. In this talk, we will introduce our continuous efforts to develop explainable decision support systems, where the system can not only tell us what to do, but also why to do so. In particular, we will introduce our recent efforts on neural logic reasoning and causal reasoning for explainable decision making, as well as their application in search engine, recommender systems, and conversational intelligent agents.