Meta-analysis of metaheuristics: Quantifying the effect of adaptiveness in adaptive large neighborhood search

Abstract

Research on metaheuristics has focused almost exclusively on (novel) algorithmic development and on competitive testing, both of which have been frequently argued to yield very little generalizable knowledge. The main goal of this paper is to promote meta-analysis — a systematic statistical examination that combines the results of several independent studies —as a more suitable way to obtain problem- and implementation-independent insights on metaheuristics. Meta-analysis is widely used in several scientific domains, most notably the medical sciences (e.g., to establish the efficacy of a certain treatment). To the best of our knowledge this is the first meta-analysis in the field of metaheuristics. To illustrate the approach, we carry out a meta-analysis to gain insights into the importance of the adaptive layer in adaptive large neighborhood search (ALNS). Although ALNS has been widely used to solve a broad range of problems, it has not yet been established whether or not adaptiveness actually contributes to the performance of an ALNS algorithm. A total of 134 studies were identified through Google Scholar or personal email correspondence with researchers in the domain, 63 of which fit our eligibility criteria. After sending requests for data to the authors of the eligible studies, we obtained results for 25 different implementations of ALNS, which were analysed using a random-effects model. On average, the addition of an adaptive layer in an ALNS algorithm improves the objective function value by 0.14% (95% confidence interval 0.07 to 0.22%). Although the adaptive layer can (and in a limited number of studies does) have an added value, it also adds considerable complexity and can therefore only be recommended in some very specific situations. These findings underline the importance of evaluating the contribution of metaheuristic components, and of knowledge over competitive testing.