Serial Correlation Test and Engineering Application of Partial Functional Linear Variable Coefficient Model Based on Quantile
Lishang Niu, Jiangwei Zhao · 2022 Second International Conference on Artificial Intelligence and Smart Energy (ICAIS) · 2022
Checking whether the residual error of the regression model has serial correlation and heteroscedasticity has always been an important work in economic and financial data analysis. In regression models, it is generally assumed that the error terms are independent of each other and have the same variance. If the assumption of independence of noise is broken. For a well-fitted model, the residuals obtained by fitting to be white noise are required, that is, the residuals no longer contain the information of the model. Therefore, the independent homoscedasticity of the error term of the model is a basic assumption. Under this assumption, regular statistical inferences can be made on the model, such as parameter estimation, hypothesis testing, etc., and further forecasts can be made. The utilization rate of engineering applications has increased by 8.3%.