Can machines learn continuously? A tutorial of the Bayesian approach

Khoat Than
Data Science Laboratory, Hanoi University of Science and Technology, Vietnam


Abstract:

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 why the Bayesian approach provides a natural and efficient answer. We will start from the basics 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, concept drifts, and stability-plasticity dilemma will be discussed.

Short bio: Khoat Than is currently Director of Data Science Laboratory, and Associate Professor at SOICT, Hanoi University of Science and Technology. 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, \u2026 His research has been being supported by various funding sources, including ONRG (US), AFOSR (US), ARL (US), NAFOSTED (VN), VINIF (VN). Some recent interests include machine learning, representation learning, deep generative models, big data.