Research on Support Vector Regression-based Profit Forecasting Model for Enterprise Economic Analysis

Abstract

Profit forecasting based on corporate economics serves as an important indicator of market returns and plays a pivotal role in market investment. This paper randomly selects financial statement data from 3,248 A-share listed companies in the H database for the years 2006–2022 as the research data source, determines the model indicator variables, and performs descriptive statistics on the sample. In constructing the prediction model, to address the limitations of traditional Support Vector Regression (SVR) models in handling large-scale data, the Whale Optimization Algorithm (WOA) was employed to optimize the penalty factor and kernel function parameters within SVR, thereby proposing an SVR-WOA-based profit prediction model. After 50 iterations, the model demonstrated a mean squared error as low as 0.075 on the validation set, exhibiting superior predictive capabilities.