A modified roach infestation optimization

Abstract

Swarm intelligence algorithms are candidate solutions to complex problems. This paper proposes a modified roach infestation optimization (MRIO) algorithm that is absolutely tied to social cockroach behaviours. MRIO improves the performance of the existing roach infestation optimization (RIO) using partial differential equation, crossover and mutation methods. The existing RIO models, made up of three components is modified and two new components are added. Simulation studies were conducted on the proposed algorithm with established benchmarks, the obtained result were compared with the results of the existing roach infestation optimization and hungry roach infestation optimization. The comparison results clearly show that the proposed algorithm outperforms the existing algorithms; and finds global optima of multi-dimensional functions.

Binary Cockroach Swarm Optimization for Combinatorial Optimization Problem

“…This algorithm was unable to solve the Rosenbrock problem optimally. Table 2 shows the results of the proposed BCSO algorithm, simulated with the dispersion model of [24]. The algorithm was able solve all of the test function problems to optimal, including the Rosenbrock function, which has been reported in the literature being too difficult to solve.…”

“…See [6,10] for details on CSO models. Two cockroach dispersion models were described in [6,24] in the literature. In this paper, the efficiency of the two models for dispersion operation was investigated through simulation.…”

“…The dispersion model of [24] was found more efficient and, hence, was used for the proposed binary version of the CSO algorithm. Table 1 shows the results of the BCSO with the dispersion model [6].…”

“…In addition, he emphasized that the general optimization terminologies, such as "solution", should be strictly used to describe any new metaheuristic [23]. The components of cockroach optimization metaheuristics are clearly described in the previous works in the literature, which include [5,6,9,10,24]. These components are constructed based on the inspiration from the social habits of natural cockroaches; each component imitates a cockroach habit; which can be adapted to solve optimization problems.…”

Development of Discrete-Cockroach Algorithm (DCA) for Feature Selection Optimization

“…However, there is still no research that tests HRIO performance to solve discrete feature selection problems with single objective optimization [10] and multi-objective optimization [11]. Obagbuwa and Adewumi [12] made modification of social cockroach behaviors, called modified roach infestation optimization (MRIO). The equations in HRIO can be modified to solve the discrete feature selection problem by substituting the algorithm for personal influence and social influence factors with the crossover and mutation methods used in GA.…”