Automatic discovery and description of human planning strategies

Discovering human planning strategies

Abstract

Our work focuses on automatic extraction of planning strategies used by people in decision-making experiments. The method we introduce uses clustering to identify similar behaviors, and then imitation learning approach from our previous work to describe those behaviors. Our tests indicate this method discovers over 50% strategies found by people 6 times faster, and could save a lot of valuable research time.

Publication
In Behavior Research Methods
Julian Skirzyński
Julian Skirzyński
PhD Candidate