Abstract：

This study focuses on machine learning-based quantitative financial portfolio optimization strategies. It first introduces the development of modern portfolio theory and the limitations of traditional methods, explaining the background of integrating machine learning techniques into portfolio optimization. By analyzing the application principles of various machine learning algorithms such as random forest, support vector machine, and neural network in portfolio optimization, we construct corresponding models and conduct empirical research using historical financial data. Results demonstrate that machine learning-based portfolio optimization strategies outperform traditional approaches in risk-adjusted returns and other metrics, providing investors with more effective asset allocation methods. The study also offers insights for future research directions.

Key Words：

machine learning; quantitative finance; portfolio optimization; risk return