Advantages Of Exchange Market Algorithm

Hybrid Exchange Market-Genetic algorithm

Names

Faculty of Electrical and Computer Engineering

University of Tabriz

Abstract−In this paper, we proposed a new hybrid algorithm that is a combination of two operational and effective algorithms. Genetic algorithm (GA) and exchange market algorithm (EMA) are two evolutionary algorithms in which they used to find the optimal answers of the different functions. EMA follows the rules in the stock market and the shareholders buy and sell. because of GA's strength in surveying solution space, it can be so helpful to combine GA with a proper exploitation based algorithm. EMA is an optimization algorithm that can truly find the global optimum of the object functions. According to the trade's inherent situation,

With comparing the results of our algorithm with the results of the eight useful and prevalent algorithms. This comparison presents the superiority of exchange market-genetic algorithm (EMGA) against other usual optimization methods.

Keywords: Genetic algorithm; Optimization; hybrid; evolutionary algorithm; Exchange market algorithm; stock market

1. Introduction

Optimization methods have become more well-liked in the last years and expanded to cover different areas of study. The last goal of all optimization algorithms is to balance the ability of exploitation and exploration efficiently in order to find global optimum [1]. There are two main ways for obtaining the optimum value. The first method is a mathematical technique that they are able to solve several problems. Heuristic algorithms are the next mechanism for optimizing in which they have more accuracy and fastness over the mathematical methods [2]. The heuristic algorithms can find the best answer of complex and constraints-based problems but mathematical solvers can't [3-7]. These algorithms are inspired by natural processes. So, various intelligent methods, such as improved ant colony search algorithm [8-11],

Especially EMA has two searcher and two absorbent operators for individuals to be absorbed by the selected person, which leads to creation and organization of random numbers in the best way. Obtaining answer earlier, search area selectivity and in turn the widespread optimization range, convergence to the identical solutions in each program iteration, and high performance in the global optimum finding are some good points of EMA [31]. GA is an adaptive search technique which simulates an evolutionary process like it is seen in nature based on the ideas of the selection of the mutation, fittest and crossing. GA follows the principles of Darwin's theory to find the solutions of a problem [32, 33]. The high adaptability and the generalizing feature of GA help to execute these problems by a noncomplex formation. GA has been successfully applied in different areas such as neural fuzzy network, fuzzy control, economic load dispatch, greenhouse climate control, path planning [34, 35]. For improving the effectiveness of GA, many researches implemented on combining of GA with other algorithms [36-38] such as Neural Networks, Dynamic Programing, Lin-Kernighan,