STOCK PRICE PREDICTION USING MACHINE LEARNING

Abstract

predicting stock market is one of the challenging tasks in the field of computation. Physical vs. physiological elements, rational vs. illogical conduct, investor emotions, market rumors, and other factors all play a role in the prediction. All of these factors combine to make stock values very fluctuating and difficult to forecast accurately. We look towards data analysis as a potential game-changer in this field. When all information about a company and stock market events is promptly available to all stakeholders/market participants, according to efficient market theory, the impacts of those occurrences are already incorporated in the stock price. As a result, it is stated that only the historical spot price accurately represents all other market events and may be used to predict future movements. As a consequence, we infer future trends using Machine Learning (ML) techniques on historical stock price data, using the previous stock price as the final representation of all influencing factors. Machine learning techniques can reveal previously undiscovered patterns and insights, which can subsequently be used to make accurate predictions. Using the LSTM (Long Short-Term Memory) model and the company's net growth calculation approach, we create a system for assessing and projecting a company's future development.