A Gaussian SVR-based intelligent framework for investment risk assessment in green finance

To enhance the efficiency and objectivity of investment risk assessment in green finance, this article proposes an intelligent evaluation model based on Gaussian kernel Support Vector Regression (SVR). The methodology begins by constructing a three-dimensional risk indicator system grounded in ESG (Environmental, Social, Governance) principles. Weights for each dimension are derived using the Analytic Hierarchy Process (AHP), and expert scores for 50 green industry projects are collected and normalized to generate standardized investment probability labels. These scores form three-dimensional feature vectors that serve as inputs to the SVR model. A Gaussian kernel function is employed to improve nonlinear mapping and address feature discreteness and weak inter-indicator correlation. The model is trained and validated via five-fold cross-validation to ensure robustness and generalizability. Experimental results demonstrate that the proposed model achieves the lowest Root Mean Square Error (RMSE = 0.096) compared to linear regression, tree-based regression, neural networks, and linear-kernel SVR. This confirms the effectiveness of the approach for small-sample scenarios and highlights its potential as a reliable and scalable tool for intelligent risk assessment in early-stage green finance applications.