Ensemble Optimization Based Feature Selection (Eofs) And Ensemble Deep Learning (Edl) Classifier For Stock Market Prediction

Abstract

The dynamic and complex financial markets are impacted by a large variety of social, political, and economic variables, stock market prediction has never been easy. This research proposes an advanced stock market forecasting structure by connecting the Ensemble optimization-based feature selection (EOFS) and the Ensemble Deep Learning (EDL) classifiers. The most relevant characteristics of the larger amount of financial data, the EOFS approach used the capabilities of many Ant Colony Optimization (ACO), Particle Swarm Optimization (PSO), and genetic algorithm (GA). The EDL classifier is then trained using its selected characteristics, which improves the accuracy and credibility of the overall forecasts, with the forecasts of many DEEP Wanda education models, such as the Conventional Neural Network (CNN), long short -term memory (LSTM) and Deep Neural Network (DNN). Experimental data shows traditional stock forecasting models' maximum accuracy, precision, and recall. Indicated technology is a property for financial analysts and investors as it provides significant growth in predicting stock prices and trends.