Can machines learn continuously? A tutorial of the Bayesian approach

Thời gian: 14:00 đến 15:30 Ngày 15/05/2019

Địa điểm: B4-705, VIASM

Báo cáo viên: Khoat Than, School of Information and Communication Technology, Hanoi University of Science and Technology

Tóm tắt:
How to build a machine that can continuously learn from observations in its life and make accurate inference/prediction? This is one of the central questions in Artificial Intelligence. Many challenges are present, such as the difficulty of learning from infinitely many observations (data), the dynamic nature of the environments, noisy and sparse data, the intractability of posterior inference, etc. This tutorial will discuss how the Bayesian approach provides a natural and efficient answer. We will start from the basic of Bayesian models, and then the variational Bayes method for inference. Next, we will discuss how to learn a Bayesian model from an infinite sequence of data. Some challenges such as catastrophic forgetting phenomenon, concept drifts, and overfitting will be discussed.
Bio: Khoat Than is currently Director of Data Science Laboratory, and Lecturer at Department of Information Systems, SOICT, HUST. He received Ph.D. (2013) from Japan Advanced Institute of Science and Technology. He joins the Program Committees of various leading conferences, including ICML, NIPS, IJCAI, ICLR, PAKDD, ACML, … His research has been being supported from various funding sources including ONRG (US), AFRL (US), ARL (US), NAFOSTED (VN). Some recent interests include machine learning, representation learning, stochastic optimization, large-scale modeling, big data.