Fitness Landscapes and Performance of Meta-Heuristics

Fonlupt, Cyril; Robilliard, Denis; Preux, Philippe; Talbi, El‐Ghazali

doi:10.1007/978-1-4615-5775-3_18

Cited by 31 publications

(31 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In some problems (and the TSP is one of them), there is a strong correlation between the cost of a solution and its "distance" to the optimum: in effect, the best solutions cluster together, i.e., have many similar components. This has been referred to in many different ways: "Massif Central" phenomenon [31], principle of proximate optimality [36], and replica symmetry [66]. If the problem under consideration has this property, it is not unreasonable to hope to find the true optimum using a biased sampling of S * .…”

Section: Global Optimization Of Ilsmentioning

confidence: 99%

Iterated Local Search: Framework and Applications

Lourenço

Martin

Stützle

2010

International Series in Operations Research &Amp; Management Science

371

262

View full text Add to dashboard Cite

Section: Global Optimization Of Ilsmentioning

confidence: 99%

Iterated Local Search: Framework and Applications

Lourenço

Martin

Stützle

2010

International Series in Operations Research &Amp; Management Science

371

262

View full text Add to dashboard Cite

“…In some problems (TSP is one of them), there is a strong correlation between the cost of a solution and its "distance" to the optimum: in effect, the best solutions cluster together, i.e., have many similar components. This has been referred to as: "Massif Central" phenomenon [10], principle of proximate optimality [16], and replica symmetry [33]. If the problem under consideration has this property, it is useful to attempt find the true optimum using a biased sampling of the space of locally optimal solutions.…”

Section: Learning Mlns With Ilsmentioning

confidence: 99%

Boosting learning and inference in Markov logic through metaheuristics

2009

View full text Add to dashboard Cite

Markov Logic (ML) combines Markov networks (MNs) and first-order logic by attaching weights to firstorder formulas and using these as templates for features of MNs. State-of-the-art structure learning algorithms in ML maximize the likelihood of a database by performing a greedy search in the space of structures. This can lead to suboptimal results because of the incapability of these approaches to escape local optima. Moreover, due to the combinatorially explosive space of potential candidates these methods are computationally prohibitive. We propose a novel algorithm for structure learning in ML, based on the Iterated Local Search (ILS) metaheuristic that explores the space of structures through a biased sampling of the set of local optima. We show through real-world experiments that the algorithm improves accuracy and learning time over the state-of-the-art algorithms. On the other side MAP and conditional inference for ML are hard computational tasks. This paper presents two algorithms for these tasks based on the Iterated Robust Tabu Search (IRoTS) metaheuristic. The first algorithm performs MAP inference and we show through extensive experiments that it improves over the state-of-theart algorithm in terms of solution quality and inference time. The second algorithm combines IRoTS steps with simulated annealing steps for conditional inference and we show through experiments that it is faster than the current stateof-the-art algorithm maintaining the same inference quality.

show abstract

“…One approach generates a representative sampling of the landscapes [17], i.e., it extracts a subset of solution with the same overall properties as the whole set of solutions. Other approaches consider search paths (steepest descent paths and descending random walks) and analyze their properties [8,22,26]. A path p of length l is a sequence of solutions p = [p1p2 .…”

Section: Definitionmentioning

confidence: 99%

A Visual Method for Analysis and Comparison of Search Landscapes

Volke

Zeckzer

Scheuermann

et al. 2015

Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation

View full text Add to dashboard Cite

Combinatorial optimization problems and corresponding (meta-)heuristics have received much attention in the literature. Especially, the structural or topological analysis of search landscapes is important for evaluating the applicability and the performance of search operators for a given problem. However, this analysis is often tedious and usually the focus is on one specific problem and only a few operators. We present a visual analysis method that can be applied to a wide variety of problems and search operators. The method is based on steepest descent walks and shortest distances in the search landscape. The visualization shows the search landscape as seen by the search algorithm. It supports the topological analysis as well as the comparison of search landscapes. We showcase the method by applying it to two different search operators on the TSP, the QAP, and the SMTTP. Our results show how differences between search operators manifest in the search landscapes and how conclusions about the suitability of the search operator for different optimizations can be drawn.

show abstract

Fitness Landscapes and Performance of Meta-Heuristics

Cited by 31 publications

References 27 publications

Iterated Local Search: Framework and Applications

Iterated Local Search: Framework and Applications

Boosting learning and inference in Markov logic through metaheuristics

A Visual Method for Analysis and Comparison of Search Landscapes

Contact Info

Product

Resources

About