Satisfiability Logic Analysis Via Radial Basis Function Neural Network with Artificial Bee Colony Algorithm

Abstract

Radial Basis Function Neural Network (RBFNN) is a variant of artificial neural network (ANN) paradigm, utilized in a plethora of fields of studies such as engineering, technology and science. 2 Satisfiability (2SAT) programming has been coined as a prominent logical rule that defines the identity of RBFNN. In this research, a swarm-based searching algorithm namely, the Artificial Bee Colony (ABC) will be introduced to facilitate the training of RBFNN. Worth mentioning that ABC is a new population-based metaheuristics algorithm inspired by the intelligent comportment of the honey bee hives. The optimization pattern in ABC was found fruitful in RBFNN since ABC reduces the complexity of the RBFNN in optimizing important parameters. The effectiveness of ABC in RBFNN has been examined in terms of various performance evaluations. Therefore, the simulation has proved that the ABC complied efficiently in tandem with the Radial Basis Neural Network with 2SAT according to various evaluations such as the Root Mean Square Error (RMSE), Sum of Squares Error (SSE), Mean Absolute Percentage Error (MAPE), and CPU Time. Overall, the experimental results have demonstrated the capability of ABC in enhancing the learning phase of RBFNN-2SAT as compared to the Genetic Algorithm (GA), Differential Evolution (DE) algorithm and Particle Swarm Optimization (PSO) algorithm.