Envisioning a Domain Learning Track for the IPC. Behnke, G. & Bercher, P. In Proceedings of the Workshop on the International Planning Competition (WIPC 2024), 2024. ![Envisioning a Domain Learning Track for the IPC [pdf]](https://bibbase.org/img/filetypes/pdf.svg "pdf")
Paper Slides abstract bibtex 5 downloads ICKEPS (International Competition on Knowledge Engineering for Planning and Scheduling) was created to make aware of the importance of domain engineering for planning and scheduling, but in past editions humans competed with the help of planning tools, thus encouraging the development of those tools (where humans were an integral part). We propose an IPC Domain Learning track, where learning algorithms would compete completely on their own, creating valid domain models without further use of input other than the constraints (like input plans) they base upon. We believe that this might help to establish some standard in the field of domain learning, such as a standard benchmark set, standard inputs (possibly an input language), and metrics to evaluate the learned domains against.
@InProceedings{Behnke2024LearningTrack,
author = {Gregor Behnke and Pascal Bercher},
booktitle = {Proceedings of the Workshop on the International Planning Competition (WIPC 2024)},
title = {Envisioning a Domain Learning Track for the IPC},
year = {2024},
abstract = {ICKEPS (International Competition on Knowledge Engineering for Planning and Scheduling) was created to make aware of the importance of domain engineering for planning and scheduling, but in past editions humans competed with the help of planning tools, thus encouraging the development of those tools (where humans were an integral part). We propose an IPC Domain Learning track, where learning algorithms would compete completely on their own, creating valid domain models without further use of input other than the constraints (like input plans) they base upon. We believe that this might help to establish some standard in the field of domain learning, such as a standard benchmark set, standard inputs (possibly an input language), and metrics to evaluate the learned domains against.},
url_Paper = {https://bercher.net/publications/2024/Behnke2024LearningTrack.pdf},
url_Slides = {https://bercher.net/publications/2024/Behnke2024LearningTrackSlides.pdf},
keywords = {workshop,DECRA}
}
Downloads: 5
{"_id":"4QoxKpFahAX623Zu8","bibbaseid":"behnke-bercher-envisioningadomainlearningtrackfortheipc-2024","author_short":["Behnke, G.","Bercher, P."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","author":[{"firstnames":["Gregor"],"propositions":[],"lastnames":["Behnke"],"suffixes":[]},{"firstnames":["Pascal"],"propositions":[],"lastnames":["Bercher"],"suffixes":[]}],"booktitle":"Proceedings of the Workshop on the International Planning Competition (WIPC 2024)","title":"Envisioning a Domain Learning Track for the IPC","year":"2024","abstract":"ICKEPS (International Competition on Knowledge Engineering for Planning and Scheduling) was created to make aware of the importance of domain engineering for planning and scheduling, but in past editions humans competed with the help of planning tools, thus encouraging the development of those tools (where humans were an integral part). We propose an IPC Domain Learning track, where learning algorithms would compete completely on their own, creating valid domain models without further use of input other than the constraints (like input plans) they base upon. We believe that this might help to establish some standard in the field of domain learning, such as a standard benchmark set, standard inputs (possibly an input language), and metrics to evaluate the learned domains against.","url_paper":"https://bercher.net/publications/2024/Behnke2024LearningTrack.pdf","url_slides":"https://bercher.net/publications/2024/Behnke2024LearningTrackSlides.pdf","keywords":"workshop,DECRA","bibtex":"@InProceedings{Behnke2024LearningTrack,\n author = {Gregor Behnke and Pascal Bercher},\n booktitle = {Proceedings of the Workshop on the International Planning Competition (WIPC 2024)},\n title = {Envisioning a Domain Learning Track for the IPC},\n year = {2024},\n abstract = {ICKEPS (International Competition on Knowledge Engineering for Planning and Scheduling) was created to make aware of the importance of domain engineering for planning and scheduling, but in past editions humans competed with the help of planning tools, thus encouraging the development of those tools (where humans were an integral part). We propose an IPC Domain Learning track, where learning algorithms would compete completely on their own, creating valid domain models without further use of input other than the constraints (like input plans) they base upon. We believe that this might help to establish some standard in the field of domain learning, such as a standard benchmark set, standard inputs (possibly an input language), and metrics to evaluate the learned domains against.},\n url_Paper = {https://bercher.net/publications/2024/Behnke2024LearningTrack.pdf},\n url_Slides = {https://bercher.net/publications/2024/Behnke2024LearningTrackSlides.pdf},\n keywords = {workshop,DECRA}\n}\n\n","author_short":["Behnke, G.","Bercher, P."],"key":"Behnke2024LearningTrack","id":"Behnke2024LearningTrack","bibbaseid":"behnke-bercher-envisioningadomainlearningtrackfortheipc-2024","role":"author","urls":{" paper":"https://bercher.net/publications/2024/Behnke2024LearningTrack.pdf"," slides":"https://bercher.net/publications/2024/Behnke2024LearningTrackSlides.pdf"},"keyword":["workshop","DECRA"],"metadata":{"authorlinks":{}},"downloads":5},"bibtype":"inproceedings","biburl":"https://bercher.net/bibtex/bibliography.bib","dataSources":["jTtEZEw8NJc375xGA","wYF8yPQT6a4TgShWe","bPpsmYWjffAy6QHP5"],"keywords":["workshop","decra"],"search_terms":["envisioning","domain","learning","track","ipc","behnke","bercher"],"title":"Envisioning a Domain Learning Track for the IPC","year":2024,"downloads":5}