Stock Price Prediction Using Sentimental Analysis & Deep Learning Optimizer Techniques

Abstract

Stock prices in India are extremely volatile for a selection of reasons, which include political verdict outcomes, rumors, budgetary news, community safety events, and so on. Because of its fluctuating nature, stock price prediction is complex and difficult. The proposed research intends to create a novel methodology by employing deep learning techniques to combine sentiment analysis using time-series data on conventional stock market prediction. It takes sentiments from online sources like social media, Twitter, and integrates sentimental duality to improve prediction accuracy. In this study, we examine the sentimental analysis effectiveness of GRU-Gradient Descent with Adam optimizer. We concentrate on analyzing stocks in order to gain a better grasp of market fluctuations and to help us estimate prices more correctly. When compared to GRU with AdaGrad, Adam optimizer, Adadelta, and RMSprop, GRU - Gradient Descent with Adam optimizer produces better results