Value investing with machine learning: the Japan-Australia market

Abstract

This research aims to investigate how machine learning technology can be effectively utilized to enhance the accuracy and reliability of stock market predictions and provide decision support for investors. This study focuses on the stock markets of Japan and Australia, and uses three classic machine learning models - Random Forest (RF), Linear Regression (LR), Support Vector Regression (SVR), as well as a hybrid deep learning model combining Convolutional Neural Networks (CNN) and Long Short-Term Memory Networks (LSTM) - to predict stock price fluctuations. The machine learning based stock selection strategy is constructed by using mean square error (MSE) and coefficient of determination (R²) to evaluate the prediction accuracy of the model, and combining indicators such as CAGR and total return. The findings of the research demonstrate that the Random Forest (RF) model achieves the highest level of prediction accuracy among the models evaluated, which provides important reference for model selection for different investment objectives. Based on the model prediction results, this study constructed an effective stock selection strategy, which validates the practical application value of machine learning technology in the stock market.