Evaluating Prediction of Stock Price using Machine Learning

The extrapolation of stock prices is an essential and unresolved problem in the sphere of finance because the results of an accurate forecast can produce considerable economic consequences and the nature of the markets makes the task difficult. This research aims at applying the concept of machine learning in forecasting of stock price for Google shares using historical data of the company’s stock for the last20 years. The qualitative aspect of the research is the collection of data with the use of the yfinance API, data preprocessing with the handling of missing values and removal of outliers. If further feature engineering, then the technical indicators included the simple moving averages and daily returns in order to improve on the capability of the model. Three types of machine learning models – Linear Regression, Random Forest, and Long Short-Term Memory (LSTM) Networks – were built experimentally and compared based on MAE and RMSE performance indices. Out of these, LSTM model provided better performance because it deals with temporal issues well by capturing temporal dependency and non linear trends in the data. In so doing, this research establishes the significance of state-of-the- art generous learning models in monetary prediction while stressing the efficacy of data origination and feature engineering. The results are quite informative for investors and financial analysts, as well as for improving the creation of further prediction models. Future work can also complement internal information with external variables like sentiment analysis and macroeconomic factors to improve their models.