Research on the Design of Financial Management Model Based on SOM-PNN Driven by Digital Economy

Abstract

This study proposes a novel financial risk prediction methodology by harnessing the power of self-organizing mapping (SOM) neural network and probabilistic neural network (PNN). The amalgamation of SOM and PNN's advantageous characteristics is seamlessly integrated into the algorithm posited within this paper. In order to collate and prognosticate data, the SOM network employs a two-dimensional topological framework comprising of two layers of neurons. Subsequently, the PNN model expeditiously furnishes the final classification outcomes by processing the output results obtained from the SOM model. The technique developed atop this composite model offers accelerated computation, effectively mitigates the impact of noisy samples, and significantly augments model accuracy. Finally, the effectiveness of the proposed method was demonstrated through a comprehensive financial risk analysis of listed companies from 2016 to 2020. The experimental results show that the SOM-PNN method has achieved high accuracy in predicting the financial difficulties experienced by traditional companies in the selected company samples, exceeding 85%. Especially when the sample data is insufficient, its accuracy reaches 80%, surpassing other algorithms. Statement: In the modern era, financial institutions use big data to perform background analysis and review, continuously optimize, and adjust, in order to introduce quantitative analysis methods into every link of risk management as far as possible. This allows financial institutions to quickly achieve balance in the game process of risk and income, and achieve Profit maximization in local or even more space. ACM CCS (2012) Classification: Computing methodologies → Machine learning → Machine learning approaches → Neural networks.