Synergizing Deep Belief Networks and Arithmetic Optimization for Stock Market Price Prediction: A Hybrid Approach

The expectation of stock market prices has become a vital and stimulating mission for both academic and practitioners’ in financial study. The unpredictable type of stock market represents that forecasting stock market prices is a stimulating mission. Generally, standard time-series predicting approaches dependent upon static development; therefore, evaluating stock value is a basic issue. Moreover, predicting the stock trend is an important concern because of the comprised variables. Consequently, between income and the economic downturn. However, a recent advancement in the machine learning, particularly from deep learning approaches are developed for researchers to practically utilize such approaches for predicting future stock trends dependent upon historical financial information, financial news, social media broadcast, and stock technical indicators (STIs). This article introduces an arithmetic optimization algorithm with deep belief network-based stock market price prediction (AOADBN-SMPP) model. The proposed technique focuses on the forecasting of stock prices in a long term. To accomplish this, the proposed model follows two key procedures namely prediction and factor optimization. In the first stage, utilizing the DNB model for forecasting stock prices. In the second stage, AOA can be leveraged to optimally fine tune the hyperparameters pertaining to the DBN technique and thereby boosts the classification execution. The AOA's layout is beneficial to optimally adjust the hyperparameters associated to the DBN version. The investigational outcome analysis of the proposed type is tested as well as the results are evaluated beneath specialized prospects. The comparative research reported a enhanced effectiveness of the suggested method over recent state of art approaches.