Portfolio advisory for stock market investment using deep learning

Abstract

This report presents a portfolio advisory system using deep learning and user-friendly interface which offers prediction and forecasting capabilities. The system focuses on three key components: UI/UX design, deep learning model, and performance analysis. The UI/UX design is an intuitive platform which is based on the clear structure of the navigation, calm loading page and welcoming homepage which has search options, watchlists and popular featured articles. The construction of deep learning model employs Long Short-Term Memory (LSTM) for stock price prediction, as it is based on pre-processing of extensive data and hyperparameter optimization to ensure accuracy. Performance analysis reviews the system by using the main elements, such as mean monthly return, annual volatility, and alpha analysis, and backtesting processes offers more realistic assessment of strategy reliability. Deploying deep learning with interactive design, the system aims at enabling an informed decision between stocks investors. This report includes the system's structure and forecasting abilities, as well as suggestions on how to enhance the presented system, for the purpose of better understanding the future potential and applicability of such systems in portfolio management.