Performance Analysis of Optimization Process on Adaptive Group of Ink Drop Spread Algorithm Proficiency

Abstract

Aims: Two evolutionary algorithms consist of Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) are being used for finding the best value of critical parameters in AGIDS which will affect the accuracy and efficiency of the algorithm. Background: Adaptive Group of Ink Drop Spread (AGIDS) is a powerful algorithm which was proposed in fuzzy domain based on Active Learning Method (ALM) algorithm. Objective: The effectiveness of AGIDS vs. artificial neural network and other soft-computing algorithms has been shown in classification, system modeling and regression problems. Methods: For solving a real-world problem a tradeoff should be taken between the needed accuracy and the available time and processing resources. Results: The simulation result shows that optimization approach will affect the accuracy of modelling being better, but its computation time is rather high. Conclusion: The simulation shows that AGIDS algorithm has a suitable efficacy in solving complex problems without using complex optimization algorithms. Other: The simulation shows that AGIDS algorithm has a suitable efficacy in solving complex problems without using complex optimization algorithms.