Out-of-the-Box and Custom Implementation of Metaheuristics. A Case Study: The Vehicle Routing Problem with Stochastic Demand

Abstract

Abstract. Metaheuristics are a class of effective algorithms for optimization problems. A basic implementation of a metaheuristic typically requires rather little development effort. With a significantly larger investment in the design, implementation, and fine-tuning, metaheuristics can often produce state-of-the-art results. According to the amount of development effort, we say that an implementation of a metaheuristic is either an out-of-the-box version or a custom one. The possibility of implementing metaheuristics in such a flexible way is one of the major strengths of these algorithms. Nonetheless, it also hides some possible catches. In particular, it should be noticed that results obtained with out-of-the-box implementations cannot be always generalized to custom ones, and vice versa. The goal of this analysis is to stress that these two ways of using metaheuristics are different. As a case study, we focus on the vehicle routing problem with stochastic demand and on five among the most successful metaheuristics-namely, tabu search, simulated annealing, genetic algorithms, iterated local search, and ant colony optimization. We show that the relative performance of these algorithms strongly varies whether one considers out-of-the-box implementations or custom ones, in which the parameters are accurately fine-tuned. Moreover, we underline the relevance of clearly stating the framework in which the results reported in the literature have been obtained. To this aim, we consider also an implementation of the same algorithms as described in the literature.

Parameter tuning of a local search heuristic for a vehicle routing problem with loading constraints

“…Authors [23] compare the performance of seven state-of-the-art algorithm configuration methods on different routing metaheuristics. Their findings confirm the results of [21] that the performance of the routing algorithm can be clearly improved by using parameter tuning. The results also reveal that there is no single best tuning method for routing algorithms, but that the Iterated F-Race algorithm seems to be the most robust one.…”

“…Parameter tuning is important because the value of the parameters may have a substantial impact on the efficacy of a heuristic algorithm [20]. Authors [21] compare the performance of five metaheuristics (tabu search, simulated annealing, genetic algorithm, iterated local search and ant colony optimization) with and without automated parameter tuning on a VRP with stochastic demands. The parameters from the non-tuned algorithms were randomly drawn within a given range, while the parameters from the tuned versions were obtained through an automatic configuration process based on the F-Race algorithm [22].…”

Loading constraints in vehicle routing problems: a focus on axle weight limits

“…Rasku et al (2014) compare the performance of seven state-of-the-art algorithm configuration methods on different routing metaheuristics. Their findings confirm the results of Pellegrini and Birattari (2011) that the performance of the routing algorithm can be clearly improved by using parameter tuning. The results also reveal that there is no single best tuning method for routing algorithms, but that the Iterated F-Race algorithm seems to be the most robust.…”

Automated Algorithm Configuration and Parameter Tuning

“…Balaprakash et al ( 2007) describe applications of F-Race, Sampling F-Race and Iterative F-Race to three high-performance stochastic local search algorithms: MAX-MIN Ant System for the TSP with 6 parameters (Stützle and Hoos, 2000), an estimation-based local search algorithm for the probabilistic TSP (PTSP) with 3 parameters (Balaprakash et al, 2010), and a simulated annealing algorithm for vehicle routing with stochastic demands (VRP-SD) with 4 parameters (Pellegrini and Birattari, 2006). The empirical results from these case studies indicate that both, Sampling F-Race and Iterative F-Race can find good configurations in spaces that are too big be handled effectively by F-Race, and that Iterative F-Race tends to give better results than Sampling F-Race, especially when applied to more difficult configuration problems.…”