Inducing Causal and Social Theories: A Prerequisite for Explanation-based Learning [chapter]

Michael J. Pazzani
1987 Proceedings of the Fourth International Workshop on MACHINE LEARNING  
We present an approach to lwming to predict and explain UK outcome oi events which lies between similaritybased methods and explanation-based methods. In the approach to learning prcsentcd in this papa and implcmcntcd in a computer program called OCC~M, a theory of causality constrams tbc starch for causal hypotheses. WC present cvidcnce that pcoplc possess a theory of causality which facilitates learning causal relationships when certain spatial and temporal relationships exist bctwccn a cause and an effect.
doi:10.1016/b978-0-934613-41-5.50027-1 fatcat:ttsuregfz5bilm3rqdji23o4ca