A model used in mutual funds performance evaluation based on SVM

Abstract

The evaluation and selection of funds are a classification problem essentially. The paper builds evaluation index sets including fund companies, fund managers, risk level and the level of risk-adjusted earnings, and uses independent samples t-test, linear regression and KPCA and SVM model to evaluate and classify securities investment funds. According to the empirical analysis, the model is practicable, and using this way to choose funds is simple, straight, and practical.