The optimal classification of enterprise secondary venture finance data can improve the statistical analysis ability of the data. The classification method of enterprise secondary venture finance data based on logical regression is put forward. The self-adaptive unsupervised learning method is used to carry on the statistical fusion processing of enterprise secondary venture finance data, and the irregular spatial clustering model of enterprise secondary venture finance data distribution is constructed. The fuzzy feature reconstruction of enterprise secondary venture finance data is carried out by using phase spatial structure reorganization method, the self-association rule feature quantity of enterprise secondary venture finance data is extracted, and the fusion clustering processing of enterprise secondary venture finance data classification is carried out by using logical regression analysis method. Combined with fuzzy C-means clustering method, the adaptive optimization and convergence control of enterprise secondary venture finance data classification are realized. To realize the classification and optimization of the secondary entrepreneurial financial data of enterprises, the simulation results show that the method has the advantages of high accuracy, good convergence and strong feature clustering ability.