Deep Learning in Financial Markets - A CNN-LSTM Approach to Stock Price Forecasting

This study investigates the use of deep learning methods in financial markets, with a specific emphasis on predicting stock prices. By employing a hybrid model combining CNN and LSTM techniques, the research delves into the complex connections present in stock price time series data. The suggested model combines CNN methods to efficiently gather features from the data of the past 10 days, along with the LSTM algorithm to understand temporal changes. Comparative assessments indicate that the CNN-LSTM technique performs better than conventional models like MLP, CNN, RNN, LSTM., and CNN-RNN in predicting stock prices in terms of reliability and accuracy. The research provides important findings for predicting future stock prices and offers a useful tool for academics and investors who are working with financial data over time. The use of Python-based TensorFlow to implement the CNN-LSTM model shows promise in uncovering valuable patterns in past stock data. The outcomes indicate its efficiency using criteria such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R-squared (R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>), demonstrating better results than older and newer models. The study advances the conversation about incorporating sophisticated deep learning methods in financial analysis, laying the groundwork for more investigation and improvement of predictive models in the ever-changing financial markets.