Learn More
Heuristic search with reachability-based heuristics is arguably the most successful paradigm in Automated Planning to date. In its earlier stages of development , heuristic search was proposed as both forward and backward search. Due to the disadvantages of backward search, in the last decade researchers focused mainly on forward search, and backward search(More)
Current evaluation functions for heuristic planning are expensive to compute. In numerous planning problems these functions provide good guidance to the solution, so they are worth the expense. However, when evaluation functions are misguiding or when planning problems are large enough, lots of node evaluations must be computed, which severely limits the(More)
This short paper provides a high-level description of the planner CBP (Cost-Based Planner). CBP performs heuristic search in the state space using several heuristics. On one hand it uses look-ahead states based on relaxed plans to speed-up the search; on the other hand the search is also guided using a numerical heuristic and a selection of actions(More)
Automatically acquiring control-knowledge for planning, as it is the case for Machine Learning in general, strongly depends on the training examples. In the case of planning, examples are usually extracted from the search tree generated when solving problems. Therefore, examples depend on the problems used for training. Traditionally, these problems are(More)
This paper focuses on heuristic cost-based planning. We propose a combination of a heuristic designed to deal with this planning model together with the usage of look-ahead states based on relaxed plans to speed-up the search. The search algorithm is a modified Best-First Search (BFS) performing Branch and Bound (B&B) to improve the last solution found. The(More)
Currently a standard technique to compute the heuris-tic in heuristic planning is to expand a planning graph on the relaxed problem. This paper presents a new approach to expand the planning graph, such that heuris-tic estimations are more accurate when an optimization metric criteria is given. Additionally, a new kind of Hill-Climbing search that combines(More)
AI planning solves the problem of generating a correct and efficient ordered set of instantiated activities, from a knowledge base of generic actions, which when executed will transform some initial state into some desirable end-state. There is a long tradition of work in AI for developing planners which make use of heuristics which are shown to improve(More)
Most of the heuristic search based planning systems perform guided search evaluating states to compute a heuristic measure. Although recent planners are quite efficient, the time spent in computing the heuristic measure is still an issue that the community has to address. In this work we present an extension to the heuristic of the relaxed plan introduced(More)
Active learning consists on the incremental generation of " good " training examples for machine learning techniques. Usually, the objective is to balance between cost of generating and analyzing all the instance space, and cost of generation of " good " examples. While there has been some work on its application to induc-tive learning techniques, there(More)