Affiliation:
1. Taylor’s Business School, Taylor’s University , Selangor , Kuala Lumpur, , Malaysia .
Abstract
Abstract
This paper proposes four major characteristic trends for financial development in the era of the digital economy, utilizing provincial data, which are specifically characterized by the development of the provincial average of the total digital financial index and the three secondary indicators of the real economy, virtual economy, and coordinated development of the real and virtual economy and incorporated with the Kernel density estimation method to enhance the overall development level of digital finance. According to the risk assessment of big data finance, the comprehensive pressure index of digital financial risk is screened, and the data financial risk early warning model is constructed by combining the random forest algorithm and the kernel principal component analysis method. Financial data characteristic variables are extracted using the nuclear principal component analysis method, and the risk level is set to predict the risk of digital financial development. Combined with provincial data, it is obtained that a mean growth of 37.20% was realized in 2015 compared with 2014, and digital finance shows a rapid development trend from 2011-2021. The early warning result of the digital finance risk early warning model in 2022 is that the probability of being in the “risk” state is small, and the likelihood of maintaining the “normal” state is significant. The early warning results are valid, and the early warning model can be further developed.