An extensive work on stock price prediction using Ant Colony Optimization algorithm (ACO-SPP)

Abstract

Stock Price Prediction (SPP) is a potential fiscally rewarding arena as it involves several dependent factors and complex structure. As the stock price fluctuates during every business day, it is nontrivial to forecast the stock prices by the buyers and sellers. Numerous methods have been developed for SPP based on Machine Learning (ML) techniques. In this work, an Ant Colony Optimization (ACO) based SPP model is introduced to forecast the stock prices accurately. The efficiency of the proposed ACO-SPP model is analyzed by comparing the performance with a set of traditional ML based classifiers such as Naive Bayes (NB), Support Vector Machine (SVM), Multilayer Perceptron (MLP), Radial Basis Function (RBF), J48, RF, Classification and Regression Tree (CART) and Olex-GA. To emphasize the benefits of ACO-SPP model, it is applied to a benchmarked dataset named Dow Jones Index (DJI) dataset and three datasets from Yahoo finance on daily, weekly and monthly basis. The experimental results infer that the ACO-SPP accurately predicts stock prices than the compared methods in terms of accuracy, F-score, AUC, Discriminant Power and Youden's Index.