{"type":"ScholarlyArticle","title":"Guiding GBFS through Learned Pairwise Rankings","authors":[{"name":"Mingyu Hao","slug":"mingyu-hao","given":"Mingyu","family":"Hao","affiliation":"Australian National University , Australia"},{"name":"Felipe Trevizan","slug":"felipe-trevizan","given":"Felipe","family":"Trevizan","affiliation":"Australian National University , Australia"},{"name":"Sylvie Thiébaux","slug":"sylvie-thiebaux","given":"Sylvie","family":"Thiébaux","affiliation":"Australian National University , Australia; LAAS-CNRS , Université de Toulouse , France"},{"name":"Patrick Ferber","slug":"patrick-ferber","given":"Patrick","family":"Ferber","affiliation":"University of Basel , Switzerland; Saarland University , Germany"},{"name":"Jörg Hoffmann","slug":"jorg-hoffmann","given":"Jörg","family":"Hoffmann","affiliation":"Saarland University , Germany"}],"abstract":"We propose a new approach based on ranking to learn to guide Greedy Best-First Search (GBFS). As previous ranking approaches, ours is based on the observation that directly learning a heuristic function is overly restrictive, and that GBFS is capable of efficiently finding good plans for a much more flexible class of total quasi-orders over states. In order to learn an optimal ranking function, we introduce a new ranking framework capable of leveraging any neural network regression model and efficiently handling the training data through batching. Compared with previous ranking approaches for planning, ours does not require complex loss functions and allows training on states outside the optimal plan with minimal overhead. Our experiments on the domains of the latest planning competition learning track show that our approach substantially improves the coverage of the underlying neural network models without degrading plan quality.","doi":"10.24963/ijcai.2024/743","url":"https://ijcai.org/proceedings/2024/743","pdf":"https://ijcai.org/proceedings/2024/0743.pdf","markdown":"https://ijcai.org/proceedings/2024/743.md","bibtex":"https://ijcai.org/proceedings/2024/bibtex/743","volume":{"slug":"2024","title":"Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2024","url":"https://ijcai.org/proceedings/2024/","year":2024,"location":"Jeju, Korea","dates":"3-9 August 2024","editors":"Kate Larson","isbn":"978-1-956792-04-1","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2024"},"section":"Planning and Scheduling","track":"Main Track","number":743,"pages":{"first":6724,"last":6732},"keywords":["Planning and Scheduling: PS: Learning in planning and scheduling"],"videos":[],"references":[{"position":1,"raw":"[Arfaee et al., 2011] Shahab Jabbari Arfaee, Sandra Zilles, and Robert C. Holte. Learning Heuristic Functions for Large State Spaces. Artificial Intelligence, 175(16- 17):2075-2098, 2011.","title":"Learning Heuristic Functions for Large State Spaces","authors":"Shahab Jabbari Arfaee, Sandra Zilles, Robert C Holte","year":"2011","venue":"Artificial Intelligence","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Bonet et al., 1997] Blai Bonet, Gábor Loerincs, and Héctor Geffner. A Robust and Fast Action Selection Mechanism for Planning. In Proc. National Conference on Artificial Intelligence (AAAI), pages 714-719, 1997.","title":"A Robust and Fast Action Selection Mechanism for Planning","authors":"Blai Bonet, Gábor Loerincs, Héctor Geffner","year":"1997","venue":"Proc. National Conference on Artificial Intelligence (AAAI)","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Burges et al., 2005] Chris Burges, Tal Shaked, Erin Ren- shaw, Ari Lazier, Matt Deeds, Nicole Hamilton, and Greg Hullender. Learning to Rank Using Gradient Descent. In Proc. International Conference on Machine learning (ICML), pages 89-96, 2005.","title":"Learning to Rank Using Gradient Descent","authors":"Chris Burges, Tal Shaked, Erin Renshaw, Ari Lazier, Matt Deeds, Nicole Hamilton, Greg Hullender","year":"2005","venue":"Proc. International Conference on Machine learning (ICML)","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Burges, 2010] Chris J.C. Burges. From RankNet to Lamb- daRank to LambdaMART: An Overview. Technical Re- port MSR-TR-2010-82, Microsoft Research, June 2010.","title":"From RankNet to Lamb-daRank to LambdaMART: An Overview","authors":"J C Chris, Burges","year":"2010","venue":null,"doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Chen et al., 2024] Dillon Z. Chen, Sylvie Thiébaux, and Fe- lipe Trevizan. Learning Domain-Independent Heuristics for Grounded and Lifted Planning. Proc. AAAI Conference on Artificial Intelligence (AAAI), 38(18):20078-20086, March 2024.","title":"Learning Domain-Independent Heuristics for Grounded and Lifted Planning","authors":"Dillon Z Chen, Sylvie Thiébaux, Felipe Trevizan","year":"2024","venue":"Proc. AAAI Conference on Artificial Intelligence (AAAI)","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Chrestien et al., 2023] Leah Chrestien, Stefan Edelkamp, Antonin Komenda, and Tomas Pevny. Optimize Plan- ning Heuristics to Rank, Not to Estimate Cost-to-Goal. In Proc. Advances in Neural Information Processing Systems (NeurIPS), volume 36, page 25508-25527, 2023.","title":"Optimize Planning Heuristics to Rank, Not to Estimate Cost-to-Goal","authors":"Leah Chrestien, Stefan Edelkamp, Antonin Komenda, Tomas Pevny","year":"2023","venue":"Proc. Advances in Neural Information Processing Systems (NeurIPS)","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Cohen and Beck, 2018] Eldan Cohen and J. Christopher Beck. Local Minima, Heavy Tails, and Search Effort for GBFS. In Proc. International Joint Conference on Artifi- cial Intelligence (IJCAI), pages 4708-4714, 2018.","title":"Local Minima, Heavy Tails, and Search Effort for GBFS","authors":"Eldan Cohen, J Christopher Beck","year":"2018","venue":"Proc. International Joint Conference on Artificial Intelligence (IJCAI)","doi":null,"url":null,"internal_url":"/proceedings/2018/654"},{"position":8,"raw":"[Damke and Hüllermeier, 2021] Clemens Damke and Eyke Hüllermeier. Ranking Structured Objects with Graph Neu- ral Networks. In Proc. International Conference on Dis- covery Science (DS), pages 166-180, 2021.","title":"Clemens Damke and Eyke Hüllermeier. Ranking Structured Objects with Graph Neural Networks","authors":"Hüllermeier Damke","year":"2021","venue":"Proc. International Conference on Discovery Science (DS)","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Ferber et al., 2020] Patrick Ferber, Malte Helmert, and Jörg Hoffmann. Neural Network Heuristics for Classical Plan- ning: A Study of Hyperparameter Space. In Proc. Euro- pean Conference on Artificial Intelligence (ECAI), pages 2346-2353, 2020.","title":"Neural Network Heuristics for Classical Planning: A Study of Hyperparameter Space","authors":"Patrick Ferber, Malte Helmert, Jörg Hoffmann","year":"2020","venue":"Proc. European Conference on Artificial Intelligence (ECAI)","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Ferber et al., 2022] Patrick Ferber, Florian Geiber, Felipe Trevizan, Malte Helmert, and Jörg Hoffmann. Neural Net- work Heuristic Functions for Classical Planning: Boot- strapping and Comparison to Other Methods. Proc. Inter- national Conference on Automated Planning and Schedul- ing (ICAPS), 32(1):583-587, June 2022.","title":"Neural Network Heuristic Functions for Classical Planning: Bootstrapping and Comparison to Other Methods","authors":"Patrick Ferber, Florian Geiber, Felipe Trevizan, Malte Helmert, Jörg Hoffmann","year":"2022","venue":"Proc. International Conference on Automated Planning and Scheduling (ICAPS)","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Freund et al., 2003] Yoav Freund, Raj Iyer, Robert E. Schapire, and Yoram Singer. An Efficient Boosting Al- gorithm for Combining Preferences. Journal of Machine Learning Research, 4:933-969, 2003.","title":"An Efficient Boosting Algorithm for Combining Preferences","authors":"Yoav Freund, Raj Iyer, Robert E Schapire, Yoram Singer","year":"2003","venue":"Journal of Machine Learning Research","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Garrett et al., 2016] Caelan Reed Garrett, Leslie Pack Kael- bling, and Tomás Lozano-Pérez. Learning to Rank for Synthesizing Planning Heuristics. In Proc. International Joint Conference on Artificial Intelligence (IJCAI), pages 3089-3095, 2016.","title":"Learning to Rank for Synthesizing Planning Heuristics","authors":"Caelan Reed Garrett, Leslie Pack Kaelbling, Tomás Lozano-Pérez","year":"2016","venue":"Proc. International Joint Conference on Artificial Intelligence (IJCAI)","doi":null,"url":null,"internal_url":"/proceedings/2016/438"},{"position":13,"raw":"[Hao et al., 2024a] Mingyu Hao, Sylvie Thiébaux, and Fe- lipe Trevizan. Source Code for \"Guiding GBFS Through Learned Pairwise Rankings\". https://doi.org/10.5281/ zenodo.11107790, 2024.","title":"Source Code for \"Guiding GBFS Through Learned Pairwise Rankings","authors":"Hao","year":"2024","venue":null,"doi":"10.5281/zenodo.11107790","url":"https://doi.org/10.5281/zenodo.11107790,2024","internal_url":null},{"position":14,"raw":"[Hao et al., 2024b] Mingyu Hao, Felipe Trevizan, Sylvie Thiébaux, Parick Ferber, and Jörg Hoffmann. Learned Pairwise Rankings for Greedy Best-First Search. In Proc. ICAPS Workshop on Reliable Data-Driven Planning and Scheduling, 2024.","title":"Learned Pairwise Rankings for Greedy Best-First Search","authors":"Hao","year":"2024","venue":"Proc. ICAPS Workshop on Reliable Data-Driven Planning and Scheduling","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Heller et al., 2022] Daniel Heller, Patrick Ferber, Julian Bit- terwolf, Matthias Hein, and Jörg Hoffmann. Neural Net- work Heuristic Functions: Taking Confidence Into Ac- count. In Proc. International Symposium on Combinato- rial Search (SOCS), pages 223-228, 2022.","title":"Neural Network Heuristic Functions: Taking Confidence Into Account","authors":"Daniel Heller, Patrick Ferber, Julian Bitterwolf, Matthias Hein, Jörg Hoffmann","year":"2022","venue":"Proc. International Symposium on Combinatorial Search (SOCS)","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Helmert, 2006] Malte Helmert. The Fast Downward Plan- ning System. Journal of Artificial Intelligence Research, 26:191-246, 2006.","title":"The Fast Downward Planning System","authors":"Malte Helmert","year":"2006","venue":"Journal of Artificial Intelligence Research","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Herbrich et al., 1999] Ralf Herbrich, Thore Graepel, and Klause Obermayer. Large Margin Rank Boundaries for Ordinal Regression. In Advances in Large Margin Classi- fiers. The MIT Press, 1999.","title":"Large Margin Rank Boundaries for Ordinal Regression","authors":"Ralf Herbrich, Thore Graepel, Klause Obermayer","year":"1999","venue":"Advances in Large Margin Classifiers","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Hoffmann and Nebel, 2001] Jörg Hoffmann and Bernhard Nebel. The FF Planning System: Fast Plan Generation Through Heuristic Search. Journal of Artificial Intelli- gence Research, 14:253-302, 2001.","title":"The FF Planning System: Fast Plan Generation Through Heuristic Search","authors":"Nebel Hoffmann, Jörg Hoffmann, Bernhard Nebel","year":"2001","venue":"Journal of Artificial Intelligence Research","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Joachims, 2002] Thorsten Joachims. Optimizing Search Engines Using Clickthrough Data. In Proc. ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), pages 133-142, 2002.","title":"Optimizing Search Engines Using Clickthrough Data","authors":"Thorsten Joachims","year":"2002","venue":"Proc. ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD)","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Karia and Srivastava, 2021] Rushang Karia and Siddharth Srivastava. Learning Generalized Relational Heuristic Networks for Model-Agnostic planning. In Proc. AAAI Conference on Artificial Intelligence (AAAI), pages 8064- 8073, 2021.","title":"Rushang Karia and Siddharth Srivastava. Learning Generalized Relational Heuristic Networks for Model-Agnostic planning","authors":"Srivastava Karia","year":"2021","venue":"Proc. AAAI Conference on Artificial Intelligence (AAAI)","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Köppel et al., 2020a] Marius Köppel, Alexander Segner, Martin Wagener, Lukas Pensel, Andreas Karwath, and Stefan Kramer. Code and Supplementary Material to the Paper: Pairwise Learning to Rank by Neural Net- works Revisited: Reconstruction, Theoretical Analy- sis and Practical Performance. https://doi.org/10.5281/ zenodo.3700819, 2020.","title":"Code and Supplementary Material to the Paper: Pairwise Learning to Rank by Neural Networks Revisited: Reconstruction","authors":"Köppel","year":"2020","venue":"Theoretical Analysis and Practical Performance","doi":"10.5281/zenodo.3700819","url":"https://doi.org/10.5281/zenodo.3700819,2020","internal_url":null},{"position":22,"raw":"[Köppel et al., 2020b] Marius Köppel, Alexander Segner, Martin Wagener, Lukas Pensel, Andreas Karwath, and Stefan Kramer. Pairwise Learning to Rank by Neural Networks Revisited: Reconstruction, Theoretical Analy- sis and Practical Performance. In Proc. European Confer- ence on Machine Learning and Knowledge Discovery in Databases (ECML PKDD), pages 237-252, 2020.","title":"Pairwise Learning to Rank by Neural Networks Revisited: Reconstruction, Theoretical Analysis and Practical Performance","authors":"Köppel","year":"2020","venue":"Proc. European Confer-ence on Machine Learning and Knowledge Discovery in Databases (ECML PKDD)","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Li, 2022] Hang Li. Learning to Rank for Information Re- trieval and Natural Language Processing. Springer, 2022.","title":"Learning to Rank for Information Retrieval and Natural Language Processing","authors":"Hang Li","year":"2022","venue":null,"doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Liu and others, 2009] Tie-Yan Liu et al. Learning to Rank for Information Retrieval. Foundations and Trends® in Information Retrieval, 3(3):225-331, 2009.","title":"Learning to Rank for Information Retrieval","authors":"Others ; Tie-Yan Liu, Liu","year":"2009","venue":"Foundations and Trends® in Information Retrieval","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Richter and Westphal, 2010] Silvia Richter and Matthias Westphal. The LAMA Planner: Guiding Cost-Based Any- time Planning with Landmarks. Journal of Artificial Intel- ligence Research, 39:127-177, 2010.","title":"The LAMA Planner: Guiding Cost-Based Anytime Planning with Landmarks","authors":"Silvia Richter, Matthias Westphal","year":"2010","venue":"Journal of Artificial Intelligence Research","doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Röger and Helmert, 2010] Gabriele Röger and Malte Helmert. The More, the Merrier: Combining Heuristic Estimators for Satisficing planning. In Proc. International Conference on Automated Planning and Scheduling (ICAPS), pages 246-249, 2010.","title":"The More, the Merrier: Combining Heuristic Estimators for Satisficing planning","authors":"Helmert Röger, Gabriele Röger, Malte Helmert","year":"2010","venue":"Proc. International Conference on Automated Planning and Scheduling (ICAPS)","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Segovia and Seipp, 2023] Javier Segovia and Jendrik Seipp. Benchmarking Repository of IPC 2023 -Learning Track. https://github.com/ipc2023-learning/benchmarks, 2023.","title":"Benchmarking Repository of IPC 2023 -Learning Track","authors":"Seipp Segovia","year":"2023","venue":null,"doi":null,"url":"https://github.com/ipc2023-learning/benchmarks,2023","internal_url":null},{"position":28,"raw":"[Seipp et al., 2020] Jendrik Seipp, Thomas Keller, and Malte Helmert. Saturated Cost Partitioning for Optimal Classi- cal Planning. Journal of Artificial Intelligence Research, 67:129-167, 2020.","title":"Saturated Cost Partitioning for Optimal Classical Planning","authors":"Seipp","year":"2020","venue":"Journal of Artificial Intelligence Research","doi":null,"url":null,"internal_url":null},{"position":29,"raw":"[Shashua and Levin, 2002] Amnon Shashua and Anat Levin. Ranking with Large Margin Principle: Two Approaches. In Proc. Advances in Neural Information Processing Sys- tems (NeurIPS), pages 937-944, 2002.","title":"Ranking with Large Margin Principle: Two Approaches","authors":"Levin ; Shashua, Amnon Shashua, Anat Levin","year":"2002","venue":"Proc. Advances in Neural Information Processing Systems (NeurIPS)","doi":null,"url":null,"internal_url":null},{"position":30,"raw":"[Shen et al., 2020] William Shen, Felipe Trevizan, and Sylvie Thiébaux. Learning Domain-Independent Planning Heuristics with Hypergraph Networks. In Proc. Interna- tional Conference on Automated Planning and Scheduling (ICAPS), volume 30, pages 574-584, 2020.","title":"Learning Domain-Independent Planning Heuristics with Hypergraph Networks","authors":"William Shen, Felipe Trevizan, Sylvie Thiébaux","year":"2020","venue":"Proc. International Conference on Automated Planning and Scheduling (ICAPS)","doi":null,"url":null,"internal_url":null},{"position":31,"raw":"[Ståhlberg et al., 2022] Simon Ståhlberg, Blai Bonet, and Hector Geffner. Learning General Optimal Policies with Graph Neural Networks: Expressive Power, Transparency, and Limits. In Proc. International Conference on Au- tomated Planning and Scheduling (ICAPS), volume 32, pages 629-637, 2022.","title":"Learning General Optimal Policies with Graph Neural Networks: Expressive Power, Transparency, and Limits","authors":"Simon Ståhlberg, Blai Bonet, Hector Geffner","year":"2022","venue":"Proc. International Conference on Automated Planning and Scheduling (ICAPS)","doi":null,"url":null,"internal_url":null},{"position":32,"raw":"[Wilt and Ruml, 2016] Christopher Makoto Wilt and Wheeler Ruml. Effective Heuristics for Suboptimal Best- First Search. Journal of Artificial Intelligence Research, 57:273-306, 2016.","title":"Effective Heuristics for Suboptimal Best-First Search","authors":"Ruml ; Wilt, Makoto Christopher, Wheeler Wilt, Ruml","year":"2016","venue":"Journal of Artificial Intelligence Research","doi":null,"url":null,"internal_url":null},{"position":33,"raw":"[Wu et al., 2010] Qiang Wu, Christopher J.C. Burges, Krysta M. Svore, and Jianfeng Gao. Adapting Boosting for Information Retrieval Measures. Information Retrieval, 13:254-270, 2010.","title":"Adapting Boosting for Information Retrieval Measures","authors":"Qiang Wu, J C Christopher, Krysta M Burges, Jianfeng Svore, Gao","year":"2010","venue":"Information Retrieval","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}