Hypothesis testing - Model Selection
3 important questions on Hypothesis testing - Model Selection
What is the goal when selecting a model?
Choose the predictors x so that the systematic part of the model captures the systematic part of the real process as well as possible.
What is overfitting? And underfitting?
Overfitting: Too many variables or interactions fits the noise of this specific sample very well but does not generalise well tot the population
Underfitting: Too simple process with too few explanatory variables or interactions leave too much unexplained variance and can lead to wrong conclusions
Underfitting: Too simple process with too few explanatory variables or interactions leave too much unexplained variance and can lead to wrong conclusions
Name 4 measures for selecting models:
- A likelihood ratio test (or F-test for normal linear models)
- Adjusted R2
- AIC / BIC
- Cross-validated measures of performance (e.g., PRESS)
The question on the page originate from the summary of the following study material:
- A unique study and practice tool
- Never study anything twice again
- Get the grades you hope for
- 100% sure, 100% understanding
Remember faster, study better. Scientifically proven.
