Improving Adaptive Neuro-Fuzzy Inference System Based on a Modified Salp Swarm Algorithm Using Genetic Algorithm to Forecast Crude Oil Price

Abstract

This paper presents a novel forecasting model for crude oil price which has the largest effect on economies and countries. The proposed method depends on improving the performance of the adaptive neuro-fuzzy inference system (ANFIS) using a modified salp swarm algorithm (SSA). The SSA simulates the behaviors of salp swarm in nature during searching for food, and it has been developed as a global optimization method. However, SSA still has some limitations such as getting trapped at a local point. Therefore, this paper uses the genetic algorithm to improve the behavior of SSA. The proposed model (GA-SSA-ANFIS) aims to determine the suitable parameters for the ANFIS by using the GA-SSA algorithm since these parameters are considered as the main factor influencing the ANFIS’s prediction process. The results of the GA-SSA-ANFIS are compared to other models, including the traditional ANFIS model, ANFIS based on GA (GA-ANFIS), ANFIS based on SSA (SSA-ANFIS) ANFIS based on particle swarm optimization (PSO-ANFIS), and ANFIS based on grey wolf optimization (GWO-ANFIS). The results show the superiority and high performances of the GA-SSA-ANFIS over the other models in predicting crude oil prices.