Exploration on Portfolio Selection and Risk Prediction in Financial Markets Based on SVM Algorithm

Literature Review

With the development of the market economy, portfolio selection and risk prediction in financial markets have increasingly become the focus of scholars’ attention. Through retrospective analysis, Aouni et al. (2018) applied methods and procedures to solve problems in portfolio selection with precision, and analyzed methods for solving problems in a multi-standard context, thereby expanding the characteristic impact of portfolio theory on mean variance. Zhang et al. (2020) proposed a cost-sensitive portfolio selection method with deep reinforcement learning, and analyzed, from a theoretical standpoint, the approximate optimal reward function proposed. Finally, based on empirical evaluation of the dataset, the effectiveness and superiority of the proposed method in terms of profitability, cost sensitivity, and presentation ability were proved. Miao (2018) employed a partial linear regression model to assess financial market stock data of varying lengths and forecasted risks associated with high-tech investments by enhancing the return and covariance matrix. Numerical examples indicated that the investment model, utilizing partial linear regression for risk prediction under inequality constraints, performs more effectively when confronted with inconsistent stock data lengths. Abensur and de Carvalho (2022) proposed an a priori classification of liquidity based on bid-ask spreads and a mathematical optimization model using liquidity as a defined participation constraint, and conducted simulation experiments using almost 20 years of data from the US and Brazilian stock exchanges. The results showed that the proposed method can lead to more successful investment portfolios. In recent years, portfolio selection and risk forecasting have made significant progress. However, with the complexity of today’s financial markets, continued improvement and optimization remain necessary in portfolio selection and risk forecasting. Existing research inadequately addresses the dynamic nature and predictive accuracy challenges within the financial market.