Definition of the Subject
The ability to produce knowledge about past experiences and exploit that knowledge in an operational context in later problem-solving and planning sessions is an important attribute of any intelligent system, human- or AI-based alike. Automated Planning and Learning is the research paradigm that focuses on the development of intelligent systems and technologies that combine the ability to make decisions and generate courses of actions (i.?e., plans) with the capability to reason and produce knowledge about past experiences, future problems that the system needs to tackle, and strategies about how to tackle them.
Probably the first work that laid a formal treatment for this combination is the early planning system STRIPS [17], developed in early 1970s. The STRIPS planning system was an evidence that planning and learning are usually two pieces of an intelligent system, where the knowledge acquired via learning is used to enhance the problem-solving and...
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
HTN planners may have other kinds of knowledge artifacts as well. For example, the \( { \textsf{SHOP2} } \) planner [39] has axioms that can be used to infer conditions about the current state.
Abbreviations
- Automated planning:
-
Automated Planning is the problem of generating a sequence of actions for an initial configuration of the world that, when executed, produces a final configuration that satisfies a specified set of goal conditions.
- Planning knowledge:
-
Almost all of successful AI planning systems developed in recent years use some sort of search-control knowledge for effective planning. Sometimes the planning knowledge is specified in terms of domain-independent heuristics, i.?e., heuristics intended for use in many different planning domains, and sometimes they are specified in terms of domain-specific control knowledge (i.?e., tailored to a specific problem domain).
- Learning for problem-solving and planning:
-
Learning for planning is the process of acquiring auxiliary knowledge related to a planning problem, which can be used by a planning system to understand better the underlying planning domain and to control its search for generating plans in that domain.
Bibliography
Aha DW (2002) Plan deconfliction, repair, and authoring in EDSS. Technical report, Progress report, Naval Research Laboratory
Ai-Chang M, Bresina J, Charest L, Hsu J, Jnsson AK, Kanefsky B, Maldague P, Morris P, Rajan K, Yglesias J (2003) Mapgen planner: Mixed-initiative activity planning for the mars exploration rover mission. In: Printed Notes of ICAPS 03 System demos
Bergmann R, Wilke W (1996) On the role of abstraction in case-based reasoning. In: European Workshop on Case-Based Reasoning (EWCBR-96), pp 28–43
Borrajo D, Veloso M (1997) Lazy incremental learning of control knowledge for efficiently obtaining quality plans. Artif Intell Rev 11(1–5):371–405
Botea A, Müller M, Schaeffer J (2005) Learning partial-order macros from solutions. In: International Conference on Automated Planning and scheduling (ICAPS-05). AAAI Press, Menlo Park, pp 231–240
Burstein M, Brinn M, Cox M, Hussain T, Laddaga R, McDermott D, McDonald D, Tomlinson R (2007) An architecture and language for the integrated learning of demonstrations. In:Burstein M, Hendler J (eds) AAAI Workshop Acquiring Planning Knowledge via Demonstration. AAAI Press, Menlo Park, pp 6–11
Chien SA (1989) Using and refining simplifications: Explanation-based learning of plans in intractable domains. In: International Joint Conference on Artificial Intelligence (IJCAI-89), Detroit. Morgan Kaufmann, San Francisco, pp 590–595
Choi D, Langley P (2005) Learning teleoreactive logic programs from problem solving. In: International Conference Inductive Logic Programming (ILP-05), Bonn. Springer, New York, pp 51–68
DeJong GF, Mooney R (1986) Explanation-based learning: An alternative view. Mach Learn 1(2):145–176
Dietterich TG (2000) Hierarchical reinforcement learning with the MAXQ value function decomposition. JAIR 13:227–303
DOT (1999) An assessment of the US marine transportation system, a report to congress. Technical report, US Department of Transportation, p 103
Edelkamp S, Hoffmann J (2004) International planning competition. http://ipc.icaps-conference.org
Erol K, Hendler J, Nau DS (1996) Complexity results for hierarchical task-network planning. AMAI 18:69–93
Erol K, Nau DS, Subrahmanian VS (1995) Complexity, decidability and undecidability results for domain-independent planning. Artif Intell 76(1–2):75–88
Estlin TA, Mooney RJ (1997) Learning to improve both efficiency and quality of planning. In: IJCAI, pp 1227–1233, Nagoya, Japan. Morgan Kaufmann, San Francisco
Fern A, Yoon S, Givan R (2004) Learning domain-specific control knowledge from random walks. In: ICAPS, Whistler. AAAI Press, Menlo Park, pp 191–199
Fikes RE, Nilsson NJ (1971) STRIPS: A new approach to the application of theorem proving to problem solving. Artif Intell 2:189–208
Gerevini A, Dimopoulos Y, Haslum P, Saetti A (2006) International planning competition. http://zeus.ing.unibs.it/ipc-5/
Goldman R (2004) Adapting research planners for applications. In: ICAPS workshop on Connecting Planning Theory with Practice
Gratch J, DeJong G (1992) Composer: A probabilistic solution to the utility problem in speed-up learning. In: AAAI, San Jose. AAAI Press, Menlo Park, pp 235–240
Hebbar K, Smith SJJ, Minis I, Nau DS (1996) Plan-based evaluation of designs for microwave modules. In: Proc ASME Design Technical Conference, Irvine, August 1996
Huang Y, Kautz H, Selman B (2000) Learning declarative control rules for constraint-based planning. In: International Conference on Machine Learning (ICML-00). Morgan Kaufmann, San Francisco, pp 415–422
Ilghami O, Nau DS, Muñoz-Avila H, Aha DW (2005) Learning preconditions for planning from plan traces and HTN structure. Comput Intell 21(4):388–413
Kambhampati S (2000) Planning graph as (dynamic) CSP: Exploiting EBL, DDB and other CSP techniques in Graphplan. JAIR 12:1–34
Kambhampati S, Katukam S, Qu Y (1996) Failure driven dynamic search control for partial order planners: An explanation based approach. Artif Intell 88(1–2):253–315
Knoblock CA (1993) Generating abstraction hierarchies: An automated approach to reducing search in planning. Kluwer, Norwell
Laird J, Newell A, Rosenbloom P (1987) SOAR: An architecture for general intelligence. Artif Intell 33(1):1–67
Lanchas J, Jimenez S, Fernandez F, Borrajo D (2007) Learning action durations from executions. In: Proceedings of the ICAPS-07 Workshop on AI Planning and Learning, Providence, 22 Sep 2007
Langley P, Choi D (2006) Learning recursive control programs from problem solving. J Mach Learn Res 7:493–518
Levine G, DeJong GF (2006) Explanation-based acquisition of planning operators. In: International Conference on Automated Planning and Scheduling (ICAPS-06), The English Iave District, 6–10 June 2006. AAAI Press, Menlo Park, pp 152–161
Matthew D, Oates T, Cohen PR (2000) Learning planning operators in real-world, partially observable environments. In: AIPS, Breckenridge, 14–17 April 2000. AAAI Press, Menlo Park
Minton S (1988) Learning effective search control knowledge: An explanation-based approach. Technical Report TR CMU-CS-88-133, School of Computer Science, Carnegie Mellon University
Mitchell T, Keller R, Kedar-Ceballi S (1986) Explanation-based generalization: A unifying view. Mach Learn 1(1):47–80
Mitchell TM (1977) Version spaces: A candidate elimination approach to rule learning. In: IJCAI. AAAI Press, Cambridge, pp 305–310
Mitchell TM (1997) Machine learning. McGraw-Hill, New York
Mooney RJ (1988) Generalizing the order of operators in macro-operators. In: Machine Learning, International Conference on Machine Learning (ML-88), Ann Arbor, 12–14 June 1988. Morgan Kaufmann, San Francisco, pp 270–283
Muñoz-Avila H, Aha DW, Breslow L, Nau DS (1999) HICAP: An interactive case-based planning architecture and its application to non-combatant evacuation operations. In: AAAI/IAAI Proceedings, Orlando, 18–22 July 1999. AAAI Press, Menlo Park, pp 870–875
Muñoz-Avila H, Breslow LA, Aha DW, Nau DS (1998) Description and functionality of NEODocTA. Technical Report AIC-96-005, Naval Research Laboratory, Navy Center for Applied Research in Artificial Intelligence
Nau DS, Cao Y, Lotem A, Muñoz-Avila H (1999) SHOP: Simple hierarchical ordered planner. In: International Joint Conference on Artificial Intelligence (IJCAI-99), Stockholm, 31 July–6 Aug 1999. Morgan Kaufmann, San Francisco, pp 968–973
Nejati N, Langley P, Konik T (2006) Learning hierarchical task networks by observation. In: Proceedings of the 23rd International Conference on Machine Learning, Pittsburgh, 25–29 June 2006. ACM Press, New York
Parr R (1998) Hierarchical control and learning for Markov decision processes. University of California, Berkeley
Reddy C, Tadepalli P (1997) Learning goal-decomposition rules using exercises. In: International Conference on Machine Learning (ICML-97), Nashville, 8–12 July 1997. Morgan Kaufmann, San Francisco, pp 278–286
Reddy C, Tadepalli P (1999) Learning horn definitions: Theory and application to planning. New Gener Comput 17(1):77–98
Ruby D, Kibler DF (1991) SteppingStone: An empirical and analytic evaluation. In: AAAI, Anaheim, July 1991. Morgan Kaufmann, San Francisco, pp 527–531
Sacerdoti E (1975) The non-linear nature of plans. In: International Joint Conference on Artificial Intelligence (IJCAI-75), Tiblisi, pp 206–214
Smith SJJ, Nau DS, Throop T (1998) Computer bridge: A big win for AI planning. AI Mag 19(2):93–105
Tate A (1977) Generating project networks. In: International Joint Cinference on Artificial Intelligence (IJCAI-77), Cambridge, 22–25 Aug 1977, pp 888–893
Wang X (1994) Learning by observation and practice: AÂ framework for automatic acquisition of planning operators. In: AAAI
Wang X (1994) Learning planning operators by observation and practice. In: International Conference on AI Planning Systems (AIPS-94), Chicago, June 1994. AAAI Press, Mento Park, pp 335–340
Watkins CJCH, Dayan P (1992) Q-learning. Mach Learn 8(3–4):279–292
Xu K, Munoz-Avila H (2005) A domain-independent system for case-based task decomposition without domain theories. In: AAAI, Pittsburgh, July 2005. AAAI Press, Menlo Park, pp 234–240
Yang Q, Wu K, Jiang Y (2005) Learning actions models from plan examples with incomplete knowledge. In: International Conference on Automated Planning and Scheduling (ICAPS-05), Monterey, June 2005. AAAI Press Menlo Park, pp 241–250
Yang Q, Wu K, Jiang Y (2007) Learning action models from plan examples using weighted max-sat. Artif Intell 171(2–3):107–143
Acknowledgments
This work was supported by DARPA's Transfer Learning and Integrated Learning programs. The opinions in this paper are those of the author and do not necessarily reflect the opinions of the funders.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag
About this entry
Cite this entry
Kuter, U. (2009). Learning and Planning (Intelligent Systems). In: Meyers, R. (eds) Encyclopedia of Complexity and Systems Science. Springer, New York, NY. https://doi.org/10.1007/978-0-387-30440-3_308
Download citation
DOI: https://doi.org/10.1007/978-0-387-30440-3_308
Publisher Name: Springer, New York, NY
Print ISBN: 978-0-387-75888-6
Online ISBN: 978-0-387-30440-3
eBook Packages: Physics and AstronomyReference Module Physical and Materials ScienceReference Module Chemistry, Materials and Physics