Model Validation for Time Series Models

The basic principles of model validation, testing the residuals and the overall model, are the same as for regression analysis. The goal of this validation is to confirm that the residuals obtained are independent, normally distributed, white noise values and that the model captures a significant portion of the overall variability. The main tools for model validation are  

Tests for Normal Distribution in time series analysis, there are three common 

Tests for Independence and Homoscedasticity these two aspects are most commonly tested together using various types of scatter plots. The most common scatter plots to examine are  

In all cases, there should not be any discernible patterns in the plots. Common scatter plots are shown in Table 3.3 (see Sect. 3.3.5 for details on how to interpret 

Using the Confidence Interval for Each of the Parameters, Oi. if the confidence interval includes 0, then the parameter can be removed from the model. Ideally, a new regression analysis excluding that parameter would need to be performed and continued until there are no more parameters to remove or add. 

---