Learning parallel portfolios of algorithms

Petrik, Marek; Zilberstein, Shlomo

doi:10.1007/s10472-007-9050-9

Cited by 24 publications

(23 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such algorithm portfolios include methods that select a single algorithm on a per-instance basis [21,7,10,31,24,26], methods that make online decisions between algorithms [16,3,23], and methods that run multiple algorithms independently on one instance, either in parallel or sequentially [13,9,20,6,27,14,22,11].…”

Section: Introductionmentioning

confidence: 99%

Evaluating Component Solver Contributions to Portfolio-Based Algorithm Selectors

Hutter

Hoos

et al. 2012

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Portfolio-based methods exploit the complementary strengths of a set of algorithms and-as evidenced in recent competitions-represent the state of the art for solving many NP-hard problems, including SAT. In this work, we argue that a state-of-the-art method for constructing portfolio-based algorithm selectors, SATzilla, also gives rise to an automated method for quantifying the importance of each of a set of available solvers. We entered a substantially improved version of SATzilla to the inaugural "analysis track" of the 2011 SAT competition, and draw two main conclusions from the results that we obtained. First, automatically-constructed portfolios of sequential, non-portfolio competition entries perform substantially better than the winners of all three sequential categories. Second, and more importantly, a detailed analysis of these portfolios yields valuable insights into the nature of successful solver designs in the different categories. For example, we show that the solvers contributing most to SATzilla were often not the overall best-performing solvers, but instead solvers that exploit novel solution strategies to solve instances that would remain unsolved without them.

show abstract

Section: Introductionmentioning

confidence: 99%

Evaluating Component Solver Contributions to Portfolio-Based Algorithm Selectors

Hutter

Hoos

et al. 2012

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…We can classify algorithm portfolios in different categories depending on: 1) how the algorithms are run and 2) when the available running time is distributed among them [16,33]. Regarding the first criterion, there are three classes [16]: parallel, where all the algorithms are run concurrently in different processors [32]; interleaved on a single processor, where the algorithms are run alternatively, simulating parallelism in one processor [14]; and sequential with restart, where at each iteration, a randomly selected algorithm is executed for a fixed amount of time (every run of the same algorithm uses a different random seed) [14].…”

Section: Related Workmentioning

confidence: 99%

“…Regarding the first criterion, there are three classes [16]: parallel, where all the algorithms are run concurrently in different processors [32]; interleaved on a single processor, where the algorithms are run alternatively, simulating parallelism in one processor [14]; and sequential with restart, where at each iteration, a randomly selected algorithm is executed for a fixed amount of time (every run of the same algorithm uses a different random seed) [14]. Respect to the second criterion, the algorithm portfolios can be classified in two categories [14,33]: static, if the distribution of the available CPU time among the algorithms is fixed before the run [32,48], or dynamic, if it is done along the search process [14].…”

Section: Related Workmentioning

confidence: 99%

Algorithm portfolio based scheme for dynamic optimization problems

Calderín¹,

Masegosa²,

Pelta³

2015

IJCIS

View full text Add to dashboard Cite

Since their first appearance in 1997 in the prestigious journal Science, algorithm portfolios have become a popular approach to solve static problems. Nevertheless and despite that success, they have not received much attention in Dynamic Optimization Problems (DOPs). In this work, we aim at showing these methods as a powerful tool to solve combinatorial DOPs. To this end, we propose a new algorithm portfolio for this type of problems that incorporates a learning scheme to select, among the metaheuristics that compose it, the most appropriate solver or solvers for each problem, configuration and search stage. This method was tested over 5 binary-coded problems (dynamic variants of OneMax, Plateau, RoyalRoad, Deceptive and Knapsack) and compared versus two reference algorithms for these problems (Adaptive Hill Climbing Memetic Algorithm and Self Organized Random Immigrants Genetic Algorithm). The results showed the importance of a good design of the learning scheme, the superiority of the algorithm portfolio against the isolated version of the metaheuristics that integrate it, and the competitiveness of its performance versus the reference algorithms.

show abstract

“…There are also examples of optimal metareasoning with respect to other object-level components such as algorithm portfolios (Petrik and Zilberstein 2006), and contract algorithms (Zilberstein et al 2003). These examples illustrate that this well-defined model of bounded rationality can be implemented in practice in many domains.…”

Section: What Object-level Decision Making Architecture Is Employed?mentioning

confidence: 99%

Metareasoning and Bounded Rationality

Zilberstein¹

2011

Metareasoning

Self Cite

View full text Add to dashboard Cite

What role does metareasoning play in models of bounded rationality? We examine the various existing computational approaches to bounded rationality and divide them into three classes. Only one of these classes significantly relies on a metareasoning component. We explore the characteristics of this class of models and argue that it offers desirable properties. In fact, many of the effective approaches to bounded rationality that have been developed since the early 1980's match this particular paradigm. We conclude with some open research problems and challenges. Computational models of bounded rationalityIn the pursuit of building decision-making machines, artificial intelligence researchers often turn to theories of "rationality" in decision theory and economics. Rationality is a desired property of intelligent agents since it provides well-defined normative evaluation criteria and since it establishes formal frameworks to analyze agents (Doyle 1990;Russell and Wefald 1991). But in general, rationality requires making optimal choices with respect to one's desires and goals. As early as 1947, Herbert Simon observed that optimal decision making is impractical in complex domains since it requires one to perform intractable computations within a limited amount of time (Simon 1947;1982). Moreover, the vast computational resources required to select optimal actions often reduce the utility of the result. Simon suggested that some criterion must be used to determine that an adequate, or satisfactory, decision has been found. He used the Scottish word "satisficing," which means satisfying, to denote decision making that searches until an alternative is found that is satisfactory by the agent's aspiration level criterion.Simon's notion of satisficing has inspired much work within the social sciences and within artificial intelligence in the areas of problem solving, planning and search. In the social sciences, much of the work has focused on developing descriptive theories of human decision making (Gigerenzer 2000). These theories attempt to explain how people make decisions in the real-world, coping with complex situations, uncertainty, and limited amount of time. The answer is often based on a variety of heuristic methods that are used by people to operate effectively in these situations. Work within the AI community-which is the focus of this paper-has produced a variety of computational models that can take into account the computational cost of decision making (Dean and Boddy 1988;Horvitz 1987;Russell et al. 1993; Wellman 1990; Zilberstein 1993). The idea that the cost of decision making must be taken into account was introduced by Simon and later by the statistician Irving Good who used the term Type II Rationality to describe it (Good 1971). Good said that "when the expected time and effort taken to think and do calculations is allowed for in the costs, then one is using the principle of rationality of type II." But neither Simon nor Good presented any effective computational framework to implement "satis...

show abstract

Learning parallel portfolios of algorithms

Cited by 24 publications

References 7 publications

Evaluating Component Solver Contributions to Portfolio-Based Algorithm Selectors

Evaluating Component Solver Contributions to Portfolio-Based Algorithm Selectors

Algorithm portfolio based scheme for dynamic optimization problems

Metareasoning and Bounded Rationality

Contact Info

Product

Resources

About