Constraint-based Causal Discovery: Conflict Resolution with Answer Set Programming

Lecturer : 
Antti Hyttinen
Event type: 
HIIT seminar
Doctoral dissertation
Respondent: 
Opponent: 
Custos: 
Event time: 
2014-08-15 10:15 to 11:15
Place: 
Kumpula, Exactum B222
Description: 
Title: Constraint-based Causal Discovery: Conflict Resolution with Answer Set Programming
 
Abstract: Recent approaches to causal discovery based on Boolean satisfiability solvers have opened new opportunities to consider search spaces for causal models with both feedback cycles and unmeasured confounders. However, the available methods have so far not been able to provide a principled account of how to handle conflicting constraints that arise from statistical variability. Here we present a new approach that preserves the versatility of Boolean constraint solving and attains a high accuracy despite the presence of statistical errors. We develop a new logical encoding of (in)dependence constraints that is both well suited for the domain and  allows for faster solving. We represent this encoding in Answer Set Programming (ASP), and apply a state-of-the-art ASP solver for the optimization task. Based on different theoretical motivations, we explore a variety of methods to handle statistical errors. Our approach currently scales to cyclic latent variable models with up to seven observed variables and outperforms the available constraint-based methods in accuracy.
 
About the presenter: Dr. Antti Hyttinen is a currently a postdoctoral scholar at California Institute of Technology. He received his Ph.D. in June 2013 from University of Helsinki, Department of Computer Science. His research interests include machine learning, probabilistic models, applications of SAT-solvers, and causality.

Last updated on 11 Aug 2014 by Sotirios Tasoulis - Page created on 11 Aug 2014 by Sotirios Tasoulis