An empirical evaluation of Martins' algorithm for the multi-objective shortest path problem

Abstract

The Shortest Path Problem is a popular optimization problem in operations research due to its wide range of practical applications. In most cases a single objective is considered, while also the multi-objective case has useful applications. The algorithms by Martins is considered very efficient. This study evaluates this algorithm by comparing it to a brute force algorithm as a first step to develop evolutionary algorithms for the multi-objective case. Experiments confirm the strength of the Martins’ algorithm.