Empirical asset pricing based on network big data mining and privacy protection

For the pricing model, a solid economic foundation is the key node to improve the pricing model. This is not only true for stocks and other profitable assets, but also for other assets such as creditor’s rights. This study is mainly based on the empirical asset pricing model, constructs a stochastic equilibrium model, and uses the generalized matrix method to analyze the asset pricing model. The estimation results show that these parameters are significant at the 5% or even 1% significance level, and the estimated values of the parameters meet the economic expectations, which can be tested by over-identification of tool variables. The experimental results prove that the empirical asset pricing model in this paper can effectively improve the effect of the single feedforward neural network model, and emphasize the necessity of feature learning, especially nonlinear unsupervised feature learning, in the application of machine learning in the field of empirical finance, which enriches the relevant research in the cross field of machine learning and empirical finance and has potential practical value.