CSIS Seminar

Computational Social Science with Latents

Abstract

Computational social science has grown to answer diverse challenges with the support from data science, machine learning, and generative models. These challenges seems to have a specific problem to solve, but the true solution only achieved when we understand the latent profiles and dynamics. This talk will provides a number of computational methodologies and case studies on such understanding. For example, we understand the population latent health-care profile with a probabilistic modeling; we look into why our politicians vote on a certain bill with a deep generative model; and we regenerate a housing market agent based model to automatically calibrate the simulation to match the real world. Along these studies, we observe how the generative models from the probabilistic modeling, the deep generative neural networks, and the agent-based model can be fused to investigate the challenges that we see in our society.

Speaker Bio

Il-Chul Moon received Ph.D. in Societal Computing from Institute of Software Research, School of Computer Science, Carnegie Mellon University in 2008. He is currently an Associate Professor at the Department of Industrial and Systems Engineering, Korea Advanced Institute of Science and Technology. His research interests include the overlapping area of computer science, management, sociology, and operations research, and also command and control analyses, health-care analyses, and disaster management.