Užití Meta-interpretativního učení k učení doménové znalosti plánovače
Název práce v češtině: | Užití Meta-interpretativního učení k učení doménové znalosti plánovače |
---|---|
Název v anglickém jazyce: | Meta-interpretive learning for Domain Control Knowledge Learning |
Klíčová slova: | planning|inductive logic programming|meta-interpretative learning |
Klíčová slova anglicky: | plánování|metainterpretativní programování |
Akademický rok vypsání: | 2024/2025 |
Typ práce: | bakalářská práce |
Jazyk práce: | |
Ústav: | Katedra teoretické informatiky a matematické logiky (32-KTIML) |
Vedoucí / školitel: | Mgr. Marta Vomlelová, Ph.D. |
Řešitel: |
Zásady pro vypracování |
The overall motivation is to explore whether Meta-interpretative learning (MIL) could help to learn the Domain Control Knowledge (DCK) for planning domains.
The student should study the current state of the MIL - meta learning of logic programs. The MIL system Louise allows to learn a simple regular or context free grammar. The student should use a set of plans for a specific planning domain as a set of positive examples for MIL learning. Resulting grammar may be too general or too specific domain knowledge. The student should whether negative examples and suitable meta-rule specifications can lead to a useful DCK. The student should explore several planning domains, for example domains used in [1]. A planning problem can be viewed as a Meta-interpretive learning: application to grammatical inference https://link.springer.com/article/10.1007/s10994-013-5358-3 The idea is to analyze plans from PDDL planners and learn a grammar. You may ignore attributes at first. It is kind of feasibility study, only preliminary experiments. The overall motivation is to explore whether MIL could help to learn the Domain Control Knowledge Attributed Transition-Based Domain Control Knowledge for Domain-Independent Planning https://ieeexplore.ieee.org/document/9253709 |
Seznam odborné literatury |
[1] L. Chrpa, R. Barták, J. Vodrážka and M. Vomlelová, "Attributed Transition-Based Domain Control Knowledge for Domain-Independent Planning," in IEEE Transactions on Knowledge and Data Engineering, vol. 34, no. 9, pp. 4089-4101, 1 Sept. 2022, doi: 10.1109/TKDE.2020.3037058.
[2] S. Patsantzis and S. H. Muggleton. Meta-Interpretive Learning as Metarule Specialisation. Machine Learning, 2021 [3] S. Patsantzis and S. H. Muggleton. Top Program Construction and Reduction for Polynomial-Time Meta-interpretive Learning. Machine Learning, 2021 [4] S.H. Muggleton, D. Lin, N. Pahlavi, and A. Tamaddoni-Nezhad. Meta-interpretive learning: application to grammatical inference. Machine Learning, 94:25-49, 2014 [5] GitHub - stassa/louise: Polynomial-time Meta-Interpretive Learning: https://github.com/stassa/louise |