Incremental Learning of Relational Action Rules. Rodrigues, C., Gerard, P., Rouveirol, C., & Soldano, H. 2010 Ninth International Conference on Machine Learning and Applications, IEEE, 2010. abstract bibtex In the Relational Reinforcement learning framework, we propose an algorithm that learns an action model allowing to predict the resulting state of each action in any given situation. The system incrementally learns a set of first order rules: each time an example contradicting the current model (a counter-example) is encountered, the model is revised to preserve coherence and completeness, by using data-driven generalization and specialization mechanisms. The system is proved to converge by storing counter-examples only, and experiments on RRL benchmarks demonstrate its good performance w.r.t state of the art RRL systems.
@article{
title = {Incremental Learning of Relational Action Rules},
type = {article},
year = {2010},
identifiers = {[object Object]},
keywords = {inductive logic programming,online and incremental learning,relational reinforcement learning},
pages = {451-458},
publisher = {IEEE},
id = {1cee7c8e-e773-3005-90f9-9edfa21844c1},
created = {2014-08-28T12:01:32.000Z},
file_attached = {false},
profile_id = {e4231fb5-c07e-3bca-b80e-916e605165b8},
last_modified = {2014-08-28T12:01:34.000Z},
read = {false},
starred = {false},
authored = {true},
confirmed = {true},
hidden = {false},
abstract = {In the Relational Reinforcement learning framework, we propose an algorithm that learns an action model allowing to predict the resulting state of each action in any given situation. The system incrementally learns a set of first order rules: each time an example contradicting the current model (a counter-example) is encountered, the model is revised to preserve coherence and completeness, by using data-driven generalization and specialization mechanisms. The system is proved to converge by storing counter-examples only, and experiments on RRL benchmarks demonstrate its good performance w.r.t state of the art RRL systems.},
bibtype = {article},
author = {Rodrigues, Christophe and Gerard, Pierre and Rouveirol, Celine and Soldano, Henry},
journal = {2010 Ninth International Conference on Machine Learning and Applications}
}
Downloads: 0
{"_id":"LFZ4EmR6BNMBAzXiN","bibbaseid":"rodrigues-gerard-rouveirol-soldano-incrementallearningofrelationalactionrules-2010","downloads":0,"creationDate":"2015-08-28T16:50:11.475Z","title":"Incremental Learning of Relational Action Rules","author_short":["Rodrigues, C.","Gerard, P.","Rouveirol, C.","Soldano, H."],"year":2010,"bibtype":"article","biburl":null,"bibdata":{"title":"Incremental Learning of Relational Action Rules","type":"article","year":"2010","identifiers":"[object Object]","keywords":"inductive logic programming,online and incremental learning,relational reinforcement learning","pages":"451-458","publisher":"IEEE","id":"1cee7c8e-e773-3005-90f9-9edfa21844c1","created":"2014-08-28T12:01:32.000Z","file_attached":false,"profile_id":"e4231fb5-c07e-3bca-b80e-916e605165b8","last_modified":"2014-08-28T12:01:34.000Z","read":false,"starred":false,"authored":"true","confirmed":"true","hidden":false,"abstract":"In the Relational Reinforcement learning framework, we propose an algorithm that learns an action model allowing to predict the resulting state of each action in any given situation. The system incrementally learns a set of first order rules: each time an example contradicting the current model (a counter-example) is encountered, the model is revised to preserve coherence and completeness, by using data-driven generalization and specialization mechanisms. The system is proved to converge by storing counter-examples only, and experiments on RRL benchmarks demonstrate its good performance w.r.t state of the art RRL systems.","bibtype":"article","author":"Rodrigues, Christophe and Gerard, Pierre and Rouveirol, Celine and Soldano, Henry","journal":"2010 Ninth International Conference on Machine Learning and Applications","bibtex":"@article{\n title = {Incremental Learning of Relational Action Rules},\n type = {article},\n year = {2010},\n identifiers = {[object Object]},\n keywords = {inductive logic programming,online and incremental learning,relational reinforcement learning},\n pages = {451-458},\n publisher = {IEEE},\n id = {1cee7c8e-e773-3005-90f9-9edfa21844c1},\n created = {2014-08-28T12:01:32.000Z},\n file_attached = {false},\n profile_id = {e4231fb5-c07e-3bca-b80e-916e605165b8},\n last_modified = {2014-08-28T12:01:34.000Z},\n read = {false},\n starred = {false},\n authored = {true},\n confirmed = {true},\n hidden = {false},\n abstract = {In the Relational Reinforcement learning framework, we propose an algorithm that learns an action model allowing to predict the resulting state of each action in any given situation. The system incrementally learns a set of first order rules: each time an example contradicting the current model (a counter-example) is encountered, the model is revised to preserve coherence and completeness, by using data-driven generalization and specialization mechanisms. The system is proved to converge by storing counter-examples only, and experiments on RRL benchmarks demonstrate its good performance w.r.t state of the art RRL systems.},\n bibtype = {article},\n author = {Rodrigues, Christophe and Gerard, Pierre and Rouveirol, Celine and Soldano, Henry},\n journal = {2010 Ninth International Conference on Machine Learning and Applications}\n}","author_short":["Rodrigues, C.","Gerard, P.","Rouveirol, C.","Soldano, H."],"bibbaseid":"rodrigues-gerard-rouveirol-soldano-incrementallearningofrelationalactionrules-2010","role":"author","urls":{},"keyword":["inductive logic programming","online and incremental learning","relational reinforcement learning"],"downloads":0},"search_terms":["incremental","learning","relational","action","rules","rodrigues","gerard","rouveirol","soldano"],"keywords":["inductive logic programming","online and incremental learning","relational reinforcement learning"],"authorIDs":[]}