Comparative Study on Stock Market Prediction using Generic CNN-LSTM and Ensemble Learning

The stock market is the platform where anyone can buy and sell or trade shares of public companies, and for that predicting the stock price helps us to forecast the future value of the company shares, derivatives, and mutual funds. The stock market is a composite and volatile system, and many factors can affect its performance. To evaluate a company's financial stability and performance, fundamental analysis is used. On the other hand, for reviewing historical price and bulk data, technical analysis has been carried out to recognize tendencies and patterns. Risk management, while Investing in the stock market carries inherent risks, and to mitigate those risks, it is crucial to spread out investments and establish stop-market orders, and other techniques. The aim of this paper is to suggest deep learning techniques in order to predict the stock prices of different companies such as AAPL(Apple), BAM(Brookfield Asset Management), and UBER and using two different models such CNN(Convolutional Neural Network) in CNN the paper uses One -Dimensional CNN(1D CNN) and LSTM(Long Short-Term Memory) uses Bidirectional LSTM(BLSTM). It implements the model on the static Apple dataset without an ensemble. While in the case of BAM, the ensemble model is on the static dataset. And, in the case of UBER, dynamic dataset for which we have fetched the dataset from Yahoo Finance(yfinance).