Recently, detection of stock markets is typically acts as a prediction problem. The intrinsic volatile behavior of stock market over the world creates a predictive operation which is a promising one. Consequently, prediction as well as diffusion modeling undermines a vast range of problems in stock market prediction (SMP). The reduction in predictive error highly decreases the investment risk. This paper presents an enhanced particle swarm optimization algorithm with control parameters (ECPSO) for decision tree (DT) classifier to forecast the direction of SMP. The proposed method involves preprocessing, feature extraction and classification. The preprocessing task is carried out with the application of exponential smoothing. Later, essential features have been obtained from preprocessed dataset. Afterwards, an ECPSO-DT model is applied for forecasting the stock prices in market. The projected method detects increasing or decreasing stock prices according to previous days.The proposed method is implemented under the application of Apple (APPL) and Facebook (FB) stocks. The derived result reveals that proposed technique provides qualified result by attaining higher accuracy compared to other methods.
Volume 12 | 05-Special Issue
Pages: 1432-1442
DOI: 10.5373/JARDCS/V12SP5/20201903