Application of Multi-Label Classification Algorithm Based on Embedded Feature Extraction in Financial Management Optimization System

The enterprise centralized financial management information system is designed and implemented. Through the integrated format and centralized accounting of financial data, managers can query and statistically analyze financial information in real time on large data screens and mobile terminals. The corresponding financial index is constructed by feature engineering, and the required data is analyzed by embedded multi-feature extraction method. The corresponding risk classification and classification modules are constructed to realize the early warning of danger. A multi-label recognition method is presented and a multi-label learning algorithm based on non-negative matrix is proposed. The influence of redundant, irrelevant and high dimensional features on multi-label classification is removed based on feature extraction of multi-label samples. Comparative trials were conducted through four published international standard databases. Experimental results show that this method can reduce the dimension of samples well. This algorithm is superior to the traditional algorithm in precision, accuracy and F-measure. Through the establishment and application of the centralized financial management system of enterprises, the value of the financial information of enterprises has been brought into full play to achieve the informationization of enterprise management. The system provides a scientific basis for enterprise risk decision.