Early warning model research of state-owned enterprises' human capital risks based on improved neural network

Abstract

With the global economic integration and development of market economy, as well as relevant national regulations and policies to support independent innovation of enterprises, human capital has become the focus of enterprises, while in recent years state-owned enterprises' flow of human capital is frequent, especially in the northeast old industrial bases, the state-owned enterprises' flow of human capital is very serious. Therefore, it is necessary to adopt BP neural network algorithm to establish a three-layer BP neural network model to evaluate the stability of human capital in state-owned enterprises, which includes 31 nodes in input layer, 22 nodes in hidden layer, and 5 nodes in output layer. When we know the stability of human capital, we can take effective measures to ensure the healthy and sustained development of state-owned enterprises.