Stock price prediction and forecasting with machine learning

Abstract

This project titled “STOCK PRICE PREDICTION AND FORECASTING WITH MACHINE LEARNING” which predict the price of stock for Bangladeshi data. Various methods such as analysis, time analysis and statistical analysis are used in estimating the market value. However, none of these methods have consistently proven to be reliable tools. Machine learning, a process that enables machines to learn from interactions in the world and expand from examples without explicit instruction, has the potential to be responsible for many important applications. Linear regression (LR) is an important technique in machine learning that can identify linear variables. Support Vector Machines (SVMs), on the other hand, have advanced features such as accuracy and prediction. LSTM has neurons that can precisely categorized features more accurately. I have implied three different model for each dataset. Each model shows satisfactory performance. Among them, linear regression performs the best for every dataset with accuracy 99.78%. SVM perform second best with accuracy 95.37%. LSTM model, performs 98.8% accuracy.