Results 1 
5 of
5
Avoiding and Escaping Depressions in RealTime Heuristic Search
"... Heuristics used for solving hard realtime search problems have regions with depressions. Such regions are bounded areas of the search space in which the heuristic function is inaccurate compared to the actual cost to reach a solution. Early realtime search algorithms, like LRTA ∗ , easily become t ..."
Abstract

Cited by 10 (3 self)
 Add to MetaCart
Heuristics used for solving hard realtime search problems have regions with depressions. Such regions are bounded areas of the search space in which the heuristic function is inaccurate compared to the actual cost to reach a solution. Early realtime search algorithms, like LRTA ∗ , easily become trapped in those regions since the heuristic values of their states may need to be updated multiple times, which results in costly solutions. Stateoftheart realtime search algorithms, like LSSLRTA ∗ or LRTA ∗ (k), improve LRTA ∗ ’s mechanism to update the heuristic, resulting in improved performance. Those algorithms, however, do not guide search towards avoiding depressed regions. This paper presents depression avoidance, a simple realtime search principle to guide search towards avoiding states that have been marked as part of a heuristic depression. We propose two ways in which depression avoidance can be implemented: markandavoid and movetoborder. We implement these strategies on top of LSSLRTA ∗ and RTAA ∗ , producing 4 new realtime heuristic search algorithms: aLSSLRTA ∗ , daLSSLRTA ∗ , aRTAA ∗ , and daRTAA ∗. When the objective is to find a single solution by running the realtime search algorithm once, we show that daLSSLRTA ∗ and daRTAA ∗ outperform their predecessors sometimes by one order of magnitude. Of the four new algorithms, daRTAA ∗ produces the best solutions given a fixed deadline on the average time allowed per planning episode. We prove all our algorithms have good theoretical properties: in finite search spaces, they find a solution if one exists, and converge to an optimal after a number of trials. 1.
Costbased heuristic search is sensitive to the ratio of operator costs
 In Proceedings of the Forth Annual Symposium on Combinatorial Search, SOCS 2012
, 2011
"... Abstract In many domains, different actions have different costs. In this paper, we show that various kinds of bestfirst search algorithms are sensitive to the ratio between the lowest and highest operator costs. First, we take common benchmark domains and show that when we increase the ratio of o ..."
Abstract

Cited by 5 (1 self)
 Add to MetaCart
Abstract In many domains, different actions have different costs. In this paper, we show that various kinds of bestfirst search algorithms are sensitive to the ratio between the lowest and highest operator costs. First, we take common benchmark domains and show that when we increase the ratio of operator costs, the number of node expansions required to find a solution increases. Second, we provide a theoretical analysis showing one reason this phenomenon occurs. We also discuss additional domain features that can cause this increased difficulty. Third, we show that searching using distancetogo estimates can significantly ameliorate this problem. Our analysis takes an important step toward understanding algorithm performance in the presence of differing costs. This research direction will likely only grow in importance as heuristic search is deployed to solve realworld problems.
RANDOM WALK PLANNING: THEORY, PRACTICE, AND APPLICATION
, 2013
"... This thesis introduces random walk (RW) planning as a new search paradigm for satisficing planning by studying its theory, its practical relevance, and applications. We develop a theoretical framework that explains the strengths and weaknesses of random walks as a tool for heuristic search. Based o ..."
Abstract

Cited by 2 (2 self)
 Add to MetaCart
This thesis introduces random walk (RW) planning as a new search paradigm for satisficing planning by studying its theory, its practical relevance, and applications. We develop a theoretical framework that explains the strengths and weaknesses of random walks as a tool for heuristic search. Based on the theory, we propose a general framework for random walk search (RWS). We identify and experimentally study the key components of RWS and for each component, design and test practical and adaptive algorithms. We study resourceconstrained planning as an application of RWS and show that the developed techniques implemented on top of RWS greatly outperform the state of the
Evaluating diversity in classical planning
 In Proceedings ICAPS
, 2014
"... Applications that require alternative plans challenge the single solution, single quality metric assumptions upon which many classical planners are designed and evaluated. To evaluate the distinctness of alternative plans (i.e., plan sets), researchers have created diversity metrics that often mea ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
Applications that require alternative plans challenge the single solution, single quality metric assumptions upon which many classical planners are designed and evaluated. To evaluate the distinctness of alternative plans (i.e., plan sets), researchers have created diversity metrics that often measure the set difference between the actions of plans. Many approaches for generating plan sets embed the same diversity metric in a weighted evaluation function to guide the search mechanism, thus confounding the search process with its evaluation. We discover that two diversity metrics fail to distinguish similar plans from each other or to identify plans with extraneous actions, so we introduce two new diversity metrics, uniqueness and overlap, to capture these cases.
Surrogate Search As a Way to Combat Harmful Effects of Illbehaved Evaluation Functions
, 2014
"... Recently, several researchers have found that costbased satisficing search with A * often runs into problems. Although some “work arounds ” have been proposed to ameliorate the problem, there has been little concerted effort to pinpoint its origin. In this paper, we argue that the origins of this p ..."
Abstract
 Add to MetaCart
(Show Context)
Recently, several researchers have found that costbased satisficing search with A * often runs into problems. Although some “work arounds ” have been proposed to ameliorate the problem, there has been little concerted effort to pinpoint its origin. In this paper, we argue that the origins of this problem can be traced back to the fact that most planners that try to optimize cost also use costbased evaluation functions (i.e., f(n) is a cost estimate). We show that costbased evaluation functions become illbehaved whenever there is a wide variance in action costs; something that is all too common in planning domains. The general solution to this malady is what we call a surrogate search, where a surrogate evaluation function that doesn’t directly track the cost objective, and is resistant to costvariance, is used. We will discuss some compelling choices for surrogate evaluation functions that are based on size rather than cost. Of particular practical interest is a costsensitive version of sizebased evaluation function where the heuristic estimates the size of cheap paths, as it provides attractive quality vs. speed tradeoffs. 1