Browsing by Author "Rubio, Jose-Miguel"

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    A Binary Machine Learning Cuckoo Search Algorithm Improved by a Local Search Operator for the Set-Union Knapsack Problem
    (2021) Garcia, Jose; Lemus-Romani, Jose; Altimiras, Francisco; Crawford, Broderick; Soto, Ricardo; Becerra-Rozas, Marcelo; Moraga, Paola; Paz Becerra, Alex; Pena Fritz, Alvaro; Rubio, Jose-Miguel; Astorga, Gino
    Optimization techniques, specially metaheuristics, are constantly refined in order to decrease execution times, increase the quality of solutions, and address larger target cases. Hybridizing techniques are one of these strategies that are particularly noteworthy due to the breadth of applications. In this article, a hybrid algorithm is proposed that integrates the k-means algorithm to generate a binary version of the cuckoo search technique, and this is strengthened by a local search operator. The binary cuckoo search algorithm is applied to the NP-hard Set-Union Knapsack Problem. This problem has recently attracted great attention from the operational research community due to the breadth of its applications and the difficulty it presents in solving medium and large instances. Numerical experiments were conducted to gain insight into the contribution of the final results of the k-means technique and the local search operator. Furthermore, a comparison to state-of-the-art algorithms is made. The results demonstrate that the hybrid algorithm consistently produces superior results in the majority of the analyzed medium instances, and its performance is competitive, but degrades in large instances.
  • No Thumbnail Available
    Item
    A binary monkey search algorithm variation for solving the set covering problem
    (2020) Crawford, Broderick; Soto, Ricardo; Olivares, Rodrigo; Embry, Gabriel; Flores, Diego; Palma, Wenceslao; Castro, Carlos; Paredes, Fernando; Rubio, Jose-Miguel
    In complexity theory, there is a widely studied grouping of optimization problems that belongs to the non-deterministic polynomial-time hard set. One of them is the set covering problem, known as one of Karp's 21-complete problems, and it consists of finding a subset of decision variables for satisfying a set of constraints at the minimum feasible cost. However, due to the nature of the problem, this cannot be solved using traditional complete algorithms for hard instances. In this work, we present an improved binary version of the monkey search algorithm for solving the set covering problem. Originally, this approximate method was naturally inspired by the cognitive behavior of monkeys for climbing mountains.We propose a new climbing process with a better exploratory capability and a newcooperation procedure to reduce the number of unfeasible solutions. For testing this approach, we present a detailed computational results section, where we illustrate how this variation of the monkey search algorithm is capable of reaching various global optimums for a well-known instance set from the easley's OR-Library and how it outperforms many other heuristics and meta-heuristics addressed in the literature. Moreover, we add a complete statistical analysis to show the effectiveness of the proposed approach with respect to the original version.