Corpus ID: 2533071

HTN-MAKER: Learning HTNs with Minimal Additional Knowledge Engineering Required

@inproceedings{Hogg2008HTNMAKERLH,
  title={HTN-MAKER: Learning HTNs with Minimal Additional Knowledge Engineering Required},
  author={C. Hogg and Hector Mu{\~n}oz-Avila and U. Kuter},
  booktitle={AAAI},
  year={2008}
}
We describe HTN-MAKER, an algorithm for learning hierarchical planning knowledge in the form of decomposition methods for Hierarchical Task Networks (HTNs). HTN-MAKER takes as input the initial states from a set of classical planning problems in a planning domain and solutions to those problems, as well as a set of semantically-annotated tasks to be accomplished. The algorithm analyzes this semantic information in order to determine which portions of the input plans accomplish a particular task… Expand
Learning hierarchical task network domains from partially observed plan traces
TLDR
A new learning algorithm, called HTNLearn, is proposed to help acquire HTN methods and action models simultaneously from partially observed plan traces, and the experimental results show that the algorithm HTN learn is both effective and efficient. Expand
From Plan Traces to Hierarchical Task Networks Using Reinforcements : A Preliminary Report
A key challenge of automated planning is the requirement of a domain expert to provide some sort of background planning knowledge about the dynamics of the planning domain. At a minimum, classicalExpand
HTN Learning via Transfer Learning of Domain Landmarks
TLDR
A novel method for landmark inference that transfers knowledge of landmarks from previously encountered environments to new environments without requiring any traces from the new environment is presented. Expand
Learning HTN Method Preconditions and Action Models from Partial Observations
TLDR
This paper presents a formal framework and algorithms to acquire HTN planning domain knowledge, by learning the preconditions and effects of actions and precond conditions of methods, and presents the HTN-learner algorithm, which first builds constraints from given observed decomposition trees to build action models and method precondition. Expand
Refining HTN Methods via Task Insertion with Preferences
TLDR
This paper proposes a novel learning framework to refine HTN methods via task insertion with completely preserving the original methods, and introduces the notion of prioritized preference to capture the incompleteness possibility of methods. Expand
Learning Methods to Generate Good Plans: Integrating HTN Learning and Reinforcement Learning
TLDR
An integration of HTN Learning with Reinforcement Learning to both learn methods by analyzing semantic annotations on tasks and to produce estimates of the expected values of the learned methods by performing Monte Carlo updates is described. Expand
Learning Hierarchical Task Networks for Nondeterministic Planning Domains
TLDR
This paper describes how to learn Hierarchical Task Networks (HTNs) in nondeterministic planning domains, where actions may have multiple possible outcomes, and developed a new learning algorithm, called HTN-MAKERND, that exploits these properties. Expand
Automatic Generation of HTNs From PDDL
Hierarchical Task Networks, or HTNs, are a popular model in planning for representing tasks or decision processes that are organized in a hierarchy. Although HTNs are known to be at least asExpand
HTN planning: Overview, comparison, and beyond
TLDR
A framework-based approach is proposed where a basis for defining different formal models of hierarchical planning is provided, and two models that comprise a large portion of HTN planners are defined, which provide a novel and comprehensive viewpoint on a core AI planning technique. Expand
Methods with Preference from HTN Planning Instances
The hierarchical task network (HTN) planning technique is used in a growing number of real-world applications. However in many domains, such as the logistics domain, as there exist thousands ofExpand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 21 REFERENCES
Learning Partial-Order Macros from Solutions
TLDR
An automated method that learns relevant information from previous experience in a domain and uses it to solve new problem instances and introduces a heuristic technique that uses only the most promising instantiations of a selected macro for node expansion. Expand
DISTILL: Learning Domain-Specific Planners by Example
TLDR
The results show that the dsPlanners automatically learned by DISTILL compactly represent its domain-specific planning experience, thus allowing them to efficiently solve problems that have not previously been encountered. Expand
Learning hierarchical task networks by observation
TLDR
An approach that observes sequences of operators taken from expert solutions to problems and learns hierarchical task networks from them and has similarities to previous algorithms for explanation-based learning, but differs in its ability to acquire hierarchical structures and in the generality of learned conditions. Expand
Learning approximate preconditions for methods in hierarchical plans
TLDR
This paper shows that the problem with slow convergence can be circumvented by having the learner generate solution plans even before the planning domain is completely learned, and reduces the size of the training set that is needed to find correct answers to a large percentage of planning problems in the test set. Expand
Building and Refining Abstract Planning Cases by Change of Representation Language
TLDR
A new abstraction methodology and a related sound and complete learning algorithm that allows the complete change of representation language of planning cases from concrete to abstract is developed. Expand
SHOP: Simple Hierarchical Ordered Planner
TLDR
In the authors' tests, SHOP was several orders of magnitude faster man Blackbox and several times faster than TLpian, even though SHOP is coded in Lisp and the other planners are coded in C. Expand
Learning Domain-Specific Control Knowledge from Random Walks
TLDR
A system for learning domain-specific control knowledge based on viewing planning domains as very large Markov decision processes and then applying a recent variant of approximate policy iteration that is bootstrapped with a new technique based on random walks is described and evaluated. Expand
A Domain-Independent System for Case-Based Task Decomposition without Domain Theories
TLDR
This work presents DInCaD (Domain-Independent System for Case-Based Task Decomposition), a system that encompasses case retrieval, refinement, and reuse, following from the idea of reusing generalized cases to solve new problems. Expand
Generalizing the Order of Operators in Macro-Operators
TLDR
An algorithm for learning partially-ordered macro-operators which has been incorporated into the EGGS domain-independent explanation-based learning system is presented and examples from the domains of computer programming and narrative understanding are used to illustrate the performance of this system. Expand
Explanation-based generalization: A unifying view
The problem of formulating general concepts from specific training examples has long been a major focus of machine learning research. While most previous research has focused on empirical methods forExpand
...
1
2
3
...