Stock Market Index Prediction via Hybrid Inertia Factor PSO and Constriction Coefficient PSO

Abstract

Conventional statistical techniques for forecasting are constrained by the underlying seasonality, non-stationary and other factors. Increasingly over the past decade, Artificial intelligence (AI) methods including Artificial Neural network (ANN), Genetic Algorithm (GA), and Particle Swarm Optimization (PSO) etc. have been used successfully to perform predictions in financial markets and other areas. This study presents a hybrid inertia factor and constriction coefficient PSO-based methodology to deal with the Stock market index problem. We will demonstrate the superiority an applicability of the proposed approach by using Tehran Stock Exchange Index (TSEI) data and comparing the outcomes with other PSO methods such as: standard PSO, Inertia Factor PSO and Constriction Coefficient PSO. Experimental results clearly show that a hybrid PSO approach meaningfully outperforms all of the other PSO methods in terms of MAD, MSE, RMSE and MAPE Evaluation statistics also, the proposed approach can be considered as a suitable AI model for stock market index forecasting problem.