论文信息 - Learning stage transition rules with Indlog

Learning stage transition rules with Indlog

This paper describes the application of an Inductive Logic Programming system to Computer Assisted Aircraft Pilot Training. The usefulness of Machine Learning techniques for Computer Assisted Aircraft Pilot Training is assessed. An Inductive Logic Programming system called Indlog is presented and some preliminary results on applying it to the Computer Assisted Aircraft Pilot Training domain are reported. The problem divides into a) learning operators (previously done by Sammut and Michie) b) learning pre and post-conditions. Comprehensible pre and post-conditions were learned using a regime in which training were augmented until predictive accuracy reached 100%. Previous hand-coded pre and post-conditions were both incomplete and incorrect.

Rui Camacho

[1] Stephen Muggleton. Optimal Layered Learning: A PAC Approach to Incremental Sampling , 1993, ALT.

[2] Richard A. Lewis,et al. Drug design by machine learning: the use of inductive logic programming to model the structure-activity relationships of trimethoprim analogues binding to dihydrofolate reductase. , 1992, Proceedings of the National Academy of Sciences of the United States of America.

[3] Stephen Muggleton,et al. Inductive Logic Programming , 2011, Lecture Notes in Computer Science.

[4] Rui Camacho,et al. Building symbolic representations of intuitive real-time skills from performance data , 1994, Machine Intelligence 13.

[5] J. Ross Quinlan,et al. Generating Production Rules from Decision Trees , 1987, IJCAI.

[6] Donald Michie,et al. Cognitive models from subcognitive skills , 1990 .

[7] Richard Fikes,et al. STRIPS: A New Approach to the Application of Theorem Proving to Problem Solving , 1971, IJCAI.