Systemic Risk Clustering of China Internet Financial Based on t-SNE Machine Learning Algorithm

TL;DR

This paper applies t-SNE machine learning to analyze China's Internet financial systemic risk across regions, revealing risk characteristics and proposing targeted regional risk classifications.

Contribution

It introduces a novel application of t-SNE for regional systemic risk classification in China's Internet finance sector.

Findings

Identified peak and thick tail risk characteristics

Classified Internet financial systemic risks into three categories

Provided regionally targeted risk management recommendations

Abstract

With the rapid development of Internet finance, a large number of studies have shown that Internet financial platforms have different financial systemic risk characteristics when they are subject to macroeconomic shocks or fragile internal crisis. From the perspective of regional development of Internet finance, this paper uses t-SNE machine learning algorithm to obtain data mining of China's Internet finance development index involving 31 provinces and 335 cities and regions. The conclusion of the peak and thick tail characteristics, then proposed three classification risks of Internet financial systemic risk, providing more regionally targeted recommendations for the systematic risk of Internet finance.