Question

1. How is the variance affected when you add a predictor to a multiple regression model?

2. Why does multiple-regression modeling require subject-matter expertise?

3. Can overfitting occur in a model with a high coefficient of determination value? What would that mean for that model?

4. What is the process of assessing a model’s capacity to make accurate predictions?

Answer 1

How is the variance affected when you add a predictor to a multiple regression model?

Ans - When we add new predictor to regression model then it affect two ways either increase or decrease

when you add new predictor and if it is correlated with others then it increases the R square which is variance explained by model and if it is uncorrelated with all of the variable in the model it decreases the variability of the model.

Why does multiple-regression modeling require subject-matter expertise?

Ans - It needs lot of experties as multiple regression need excellent understanding of model building procedure. As it need attention to various things like Overfitting, underfitting, multicollinearity, Accuracy of the model and so on.

Can overfitting occur in a model with a high coefficient of determination value? What would that mean for that model?

Ans - YES overfitting in a model can occured with high coefficient of determnation as you add the variable in the model it always increases the coefficient of determination and hence it is required to use adjusted coefficient of determination. That indicates the model has very low predictive power.

What is the process of assessing a model’s capacity to make accurate predictions?

Ans - Following are some stages which can be used for model assesment,

1) To check assumptions

2) To check multicollinearity

3) To check area under the curve

4) To check adjusted R square

5) To check error term if error is to high then models prediction accuracy will be lower