Dennis Weyland

A Rigorous Analysis of the Harmony Search Algorithm: How the Research Community can be Misled by a Novel Methodology

In recent years a lot of novel (mostly naturally inspired) search heuristics have been proposed. Among those approaches is Harmony Search. After its introduction in 2000, positive results and improvements over existing approaches have been reported. In this paper, the authors give a review of the developments of Harmony Search during the past decade and perform a rigorous analysis of this approach. This paper compares Harmony Search to the well-known search heuristic called Evolution Strategies. Harmony Search is a special case of Evolution Strategies in which the authors give compelling evidence for the thesis that research in Harmony is fundamentally misguided. The overarching question is how such a method could be inaccurately portrayed as a significant innovation without confronting a respectable challenge of its content or credentials. The authors examine possible answers to this question, and implications for evaluating other procedures by disclosing the way in which limitations of the method have been systematically overlooked.

Coupling ant colony systems with strong local searches

Ant colony system is a well known metaheuristic framework, and many efficient algorithms for different combinatorial optimization problems have been derived from this general framework. In this paper some directions for improving the original framework when a strong local search routine is available, are identified. In particular, some modifications able to speed up the method and make it competitive on large problem instances, on which the original framework tends to be weaker, are described. The resulting framework, called Enhanced Ant Colony System is tested on three well-known combinatorial optimization problems arising in the transportation field. Many new best known solutions are retrieved for the benchmarks available for these optimization problems.

An Enhanced Ant Colony System for the Team Orienteering Problem with Time Windows

The Team Orienteering Problem with Time Windows (TOPTW) is a combinatorial optimization problem arising both in industrial scheduling and in transportation. Ant Colony System (ACS) is a well-known metaheuristic framework, and many efficient algorithms for different optimization problems - among them the TOPTW - have been derived from this general framework. In this paper some directions for improving an ACS algorithm for the TOPTW recently appeared in the literature are identified. The resulting algorithm, called Enhanced Ant Colony System is experimentally shown to be extremely effective on some well-known benchmark instances available in the literature.

Simulated annealing, its parameter settings and the longest common subsequence problem

Simulated Annealing is a probabilistic search heuristic for solving optimization problems and is used with great success on real life problems. In its standard form Simulated Annealing has two parameters, namely the initial temperature and the cooldown factor. In literature there are only rules of the thumb for choosing appropriate parameter values. This paper investigates the influence of different values for these two parameters on the optimization process from a theoretical point of view and presents some criteria for problem specific adjusting of these parameters. With these results the performance of the Simulated Annealing algorithm on solving the Longest Common Subsequence Problem is analysed using different values for the two parameters mentioned above. For all these parameter settings it is proved that even rather simple input instances of the Longest Common Subsequence Problem can neither be solved to optimality nor approximately up to an approximation factor arbitrarily close to 2 efficiently.

Analysis of evolutionary algorithms for the longest common subsequence problem

In the longest common subsequence problem the task is to find the longest sequence of letters that can be found as subsequence in all members of a given finite set of sequences. The problem is one of the fundamental problems in computer science with the task of finding a given pattern in a text as an important special case. It has applications in bioinformatics, problem-specific algorithms and facts about its complexity are known. Motivated by reports about good performance of evolutionary algorithms for some instances of this problem a theoretical analysis of a generic evolutionary algorithm is performed. The general algorithmic framework encompasses EAs as different as steady state GAs with uniform crossover and randomized hill-climbers. For all these algorithms it is proved that even rather simple special cases of the longest common subsequence problem can neither be solved to optimality nor approximately solved up to an approximation factor arbitrarily close to 2.

An Enhanced Ant Colony System for the Sequential Ordering Problem

A well-known Ant Colony System algorithm for the Sequential Ordering Problem is studied to identify its drawbacks. Some criticalities are identified, and an Enhanced Ant Colony System method that tries to overcome them, is proposed. Experimental results show that the enhanced method clearly outperforms the original algorithm and becomes a reference method for the problem under investigation.

New Approximation-Based Local Search Algorithms for the Probabilistic Traveling Salesman Problem

In this paper we present new local search algorithms for the Probabilistic Traveling Salesman Problem (PTSP) using sampling and ad-hoc approximation. These algorithms improve both runtime and solution quality of state-of-the-art local search algorithms for the PTSP.

Hardness results for the probabilistic traveling salesman problem with deadlines

The Probabilistic Traveling Salesman Problem with Deadlines (PTSPD) is a Stochastic Vehicle Routing Problem considering time dependencies. Even the evaluation of the objective function is considered to be a computationally demanding task. So far there is no evaluation method known that guarantees a polynomial runtime, but on the other hand there are also no hardness results regarding the PTSPD objective function. In our work we show that the evaluation of the objective function of the PTSPD, even for Euclidean instances, is #P-hard. In fact, we even show that computing the probabilities, with which deadlines are violated is #P-hard. Based on this result we additionally show that the decision variant of the Euclidean PTSPD, the optimization variant of the Euclidean PTSPD and delta evaluation in reasonable local search neighborhoods is #P-hard.

A metaheuristic framework for stochastic combinatorial optimization problems based on GPGPU with a case study on the probabilistic traveling salesman problem with deadlines

In this work we propose a general metaheuristic framework for solving stochastic combinatorial optimization problems based on general-purpose computing on graphics processing units (GPGPU). This framework is applied to the probabilistic traveling salesman problem with deadlines (PTSPD) as a case study. Computational studies reveal significant improvements over state-of-the-art methods for the PTSPD. Additionally, our results reveal the huge potential of the proposed framework and sampling-based methods for stochastic combinatorial optimization problems.

Heuristics for the probabilistic traveling salesman problem with deadlines based on quasi-parallel Monte Carlo sampling

The Probabilistic Traveling Salesman Problem with Deadlines (PTSPD) is a Stochastic Vehicle Routing Problem with a computationally demanding objective function. In this work we propose an approximation for that objective function based on Monte Carlo Sampling and using the novel approach of quasi-parallel evaluation of samples. We perform comprehensive computational studies that reveal the efficiency of this approximation. Additionally, we examine different Local Search Algorithms and present a Random Restart Local Search Algorithm for solving the PTSPD together with an extensive computational study on a large set of benchmark instances.