ML92 Proceedings of the ninth international workshop on Machine learning
ACM SIGART Bulletin
HTN planning: complexity and expressivity
AAAI'94 Proceedings of the twelfth national conference on Artificial intelligence (vol. 2)
Learning action strategies for planning domains
Artificial Intelligence
Explanation-Based Learning: An Alternative View
Machine Learning
Learning Goal-Decomposition Rules using Exercises
ICML '97 Proceedings of the Fourteenth International Conference on Machine Learning
Learning approximate preconditions for methods in hierarchical plans
ICML '05 Proceedings of the 22nd international conference on Machine learning
Learning hierarchical task networks by observation
ICML '06 Proceedings of the 23rd international conference on Machine learning
A goal- and dependency-directed algorithm for learning hierarchical task networks
Proceedings of the fifth international conference on Knowledge capture
HTN-MAKER: learning HTNs with minimal additional knowledge engineering required
AAAI'08 Proceedings of the 23rd national conference on Artificial intelligence - Volume 2
Journal of Artificial Intelligence Research
Learning teleoreactive logic programs from problem solving
ILP'05 Proceedings of the 15th international conference on Inductive Logic Programming
A goal- and dependency-directed algorithm for learning hierarchical task networks
Proceedings of the fifth international conference on Knowledge capture
Hi-index | 0.00 |
We describe a new approach for learning procedural knowledge represented as teleoreactive logic programs using relational behavior traces as input. This representation organizes task decomposition skills hierarchically and associate explicitly defined goals with them. Our approach integrates analytical learning with inductive generalization in order to learn these skills. The analytical component predicts the goal dependencies in a successful solution and generates a teleoreactive logic program that can solve similar problems by determining the structure of the skill hierarchy and skill applicability conditions (preconditions), which may be overgeneral. The inductive component experiments with these skills on new problems and uses the data collected in this process to refine the preconditions. Our system achieves this by converting the data collected during the problem solving experiments into the positive and negative examples of preconditions that can be learned with a standard Inductive Logic Programming system. We show that this conversion uses one of the main commitments of teleoreactive logic programs: associating all skills with explicitly defined goals. We claim that our approach uses less expert effort compared to a purely inductive approach and performs better compared to a purely analytical approach.