Binary Artificial Bee Colony Algorithms for {0-1} Advertisement Problem

Abstract

A company wants to select magazine publishers for advertising. It is vital for the company to maximize the number of subscribers reached by advertising while not exceeding its advertising budget. To determine which publishers to select can be considered as an advertisement optimization problem and can be solved by using binary metaheuristics. Artificial Bee Colony (ABC) is one of the most popular metaheuristics in the field of swarm intelligence and has been widely used in numerical optimization and engineering applications. To solve binary advertisement problem, three binary variants of ABC algorithm are proposed in this paper. The first binary variant is based on Sigmoid transfer function and indicated as sigABC, while the second and the third binary variants are based on exclusive OR (xor) and crossover genetic operators, respectively, and are indicated as xorABC and crossoverABC. The proposed binary ABC variants are evaluated using 20 instance-dataset. The comparative experimental results show the superior performance of the crossoverABC and sigABC methods compared to xorABC technique.