Acquiring domain-specific planners by example

Abstract : Intelligent problem solving requires the ability to select actions autonomously from a specific state to reach objectives. Planning algorithms provide approaches to look ahead and select a complete sequence of actions. Given a domain description consisting of preconditions and effects of the actions the planner can take, an initial state, and a goal, a planning program returns a sequence of actions to transform the initial state into a state in which the goal is satisfied. Classical planning research has addressed this problem in a domain-independent manner--the same algorithm generates a complete plan for any domain specification. This feature comes at a cost which domain-independent planners incur either in high search efforts or in tedious hand-coded domain knowledge. Previous approaches to efficient general-purpose planning have focused on reducing the search involved in an existing general-purpose planning algorithm. An interesting alternative is to use example plans in a particular domain to demonstrate how to solve problems in that domain and to use that information to solve new problems independently of a domain-independent planner. Others have used example plans for case based planning, but the retrieval and adaptation mechanisms were still domain-independent and efficiency issues were still a concern. In my thesis, I propose to introduce algorithms to extract complex, repeating processes, in the form of domain-specific planning programs, from example plans. I will investigate the application of these learned programs to modelling agent preferences and choices. I will also investigate how the programs can be used, extended, and repaired dynamically as an agent encounters new problems and acquires new experience. Finally, I will compare the template-based planning paradigm to existing general-purpose and domain-specific planning programs with a full evaluation on new and existing planning domains.

[1]  Daniel S. Weld,et al.  UCPOP: A Sound, Complete, Partial Order Planner for ADL , 1992, KR.

[2]  Douglas R. Smith,et al.  KIDS - A Knowledge-Based Software Development System , 1991 .

[3]  Manuela M. Veloso,et al.  OBDD-based Universal Planning for Synchronized Agents in Non-Deterministic Domains , 2000, J. Artif. Intell. Res..

[4]  Earl D. Sacerdott.  Planning in a hierarchy of abstraction spaces , 1973, IJCAI 1973.

[5]  David E. Smith,et al.  Conditional Effects in Graphplan , 1998, AIPS.

[6]  Yumi Iwasaki,et al.  The concept and implementation of skeletal plans , 1985, Journal of Automated Reasoning.

[7]  Henry A. Kautz,et al.  Generalized Plan Recognition , 1986, AAAI.

[8]  Mark A. Peot,et al.  Suspending Recursion in Causal-Link Planning , 1996, AIPS.

[9]  Ralph Bergmann.  Knowledge Acquisition by Generating Skeletal Plans from Real World Cases , 1991, Contemporary Knowledge Engineering and Cognition.

[10]  Charles W. Anderson,et al.  Strategy Learning with Multilayer Connectionist Representations , 1987 .

[11]  Manuela M. Veloso,et al.  Planning and Learning by Analogical Reasoning , 1994, Lecture Notes in Computer Science.

[12]  Manuela M. Veloso,et al.  Prodigy/Analogy: Analogical Reasoning in General Problem Solving , 1993, EWCBR.

[13]  Zohar Manna,et al.  Fundamentals of Deductive Program Synthesis , 1992, IEEE Trans. Software Eng..

[14]  Marcel Schoppers,et al.  Universal Plans for Reactive Robots in Unpredictable Environments , 1987, IJCAI.

[15]  Michael Anthony Bauer,et al.  Programming by Examples , 1986, Artif. Intell..

[16]  Robert S. Williams.  Learning to Program by Examining and Modifying Cases , 1988, ML.

[17]  James A. Hendler,et al.  A Validation-Structure-Based Theory of Plan Modification and Reuse , 1992, Artif. Intell..

[18]  David Leake,et al.  Case-Based Reasoning: Experiences, Lessons and Future Directions , 1996 .

[19]  Jaime G. Carbonell,et al.  Learning effective search control knowledge: an explanation-based approach , 1988 .

[20]  Daniel S. Weld.  An Introduction to Least Commitment Planning , 1994, AI Mag..

[21]  Craig A. Knoblock.  Automatically Generating Abstractions for Planning , 1994, Artif. Intell..

[22]  Richard Fikes,et al.  Learning and Executing Generalized Robot Plans , 1993, Artif. Intell..

[23]  Michael Luck,et al.  Multi-Agent Systems and Applications , 2001, Lecture Notes in Computer Science.

[24]  Manuela M. Veloso,et al.  The Lumberjack Algorithm for Learning Linked Decision Forests , 2000, PRICAI.

[25]  Mark A. Peot,et al.  Postponing Threats in Partial-Order Planning , 1993, AAAI.

[26]  Jaime G. Carbonell,et al.  Towards a General Framework for Composing Disjunctive and Iterative Macro-operators , 1989, IJCAI.

[27]  Paolo Traverso,et al.  Strong Planning in Non-Deterministic Domains Via Model Checking , 1998, AIPS.

[28]  Paolo Traverso,et al.  Automatic OBDD-Based Generation of Universal Plans in Non-Deterministic Domains , 1998, AAAI/IAAI.

[29]  Roni Khardon,et al.  Learning Action Strategies for Planning Domains , 1999, Artif. Intell..

[30]  Subbarao Kambhampati,et al.  Explanation-Based Generalization of Partially Ordered Plans , 1991, AAAI.

[31]  Richard C. Waters,et al.  Approaches to Automatic Programming , 1993, Adv. Comput..

[32]  Craig A. Knoblock.  Learning Abstraction Hierarchies for Problem Solving , 1990, AAAI.

[33]  Ute Schmid,et al.  Inductive Synthesis of Functional Programs , 2003, Lecture Notes in Computer Science.

[34]  Fritz Wysotzki,et al.  Applying Inductive Program Synthesis to Macro Learning , 2000, AIPS.

[35]  James A. Hendler,et al.  Flexible reuse and modification in hierarchical planning: a validation structure-based approach , 1989 .

[36]  Christer Bäckström.  Finding Least Constrained Plans and Optimal Parallel Executions is Harder than We Thought , 1994 .

[37]  Richard Fikes,et al.  STRIPS: A New Approach to the Application of Theorem Proving to Problem Solving , 1971, IJCAI.

[38]  Blai Bonet,et al.  The AIPS-98 Planning Competition , 2000, AI Mag..

[39]  J. Harris.  Where is the child's environment? A group socialization theory of development. , 1995 .

[40]  Steven Minton,et al.  Selectively Generalizing Plans for Problem-Solving , 1985, IJCAI.

[41]  Xuemei Wang,et al.  Learning Planning Operators by Observation and Practice , 1994, AIPS.

[42]  Manuela M. Veloso,et al.  Analyzing Plans with Conditional Effects , 2002, AIPS.

[43]  J. Shavlik.  Acquiring Recursive and Iterative Concepts with Explanation-Based Learning , 1990, Machine Learning.

[44]  M. Veloso,et al.  Nonlinear Planning with Parallel Resource Allocation , 1990 .

[45]  Edwin P. D. Pednault,et al.  FORMULATING MULTIAGENT, DYNAMIC-WORLD PROBLEMS IN THE CLASSICAL PLANNING FRAMEWORK , 1987 .

[46]  S. Kambhampati,et al.  Learning Explanation-Based Search Control Rules for Partial Order Planning , 1994, AAAI.

[47]  Robert Wilensky,et al.  A model for planning in complex situations , 1981 .

[48]  Larry D. Pyeatt,et al.  Decision Tree Function Approximation in Reinforcement Learning , 1999 .

[49]  Pierre Régnier,et al.  Complete Determination of Parallel Actions and Temporal Optimization in Linear Plans of Action , 1991, EWSP.

[50]  Thomas G. Dietterich.  What is machine learning? , 2020, Archives of Disease in Childhood.

[51]  De,et al.  Relational Reinforcement Learning , 2001, Encyclopedia of Machine Learning and Data Mining.

[52]  Richard E. Korf,et al.  Macro-Operators: A Weak Method for Learning , 1985, Artif. Intell..

[53]  Jaime G. Carbonell,et al.  Learning by experimentation: the operator refinement method , 1990 .

[54]  Kristian J. Hammond,et al.  CHEF: A Model of Case-Based Planning , 1986, AAAI.

[55]  Manuela Veloso,et al.  Automatically Acquiring Planning Templates from Example Plans , 2002 .

[56]  Luc De Raedt,et al.  Inductive Logic Programming: Theory and Methods , 1994, J. Log. Program..

[57]  Pedro M. Domingos,et al.  Programming by demonstration: a machine learning approach , 2001 .

[58]  Avrim Blum,et al.  Fast Planning Through Planning Graph Analysis , 1995, IJCAI.