Econ 524: AIC and BIC Model - Statistics Assignment

Statistics Assignment Help

Model Selection 1. For each of the five models from project 4, compute the AIC and the BIC. Which model is selected by each of these metrics? 2. For each model which is nested by another model, compute likelihood ratio test statistics and p-values. What is the preferred model using these tests? 3. Perform a 5-folds cross-validation for every model. Reset the seed to 100 before doing the cross-validation for each model. Since we are using k=5, each fold should have 4 subjects (and each fold should have the full 100 rounds for each of the 4 subjects). Report the cross-validated MSE for each of the five models. What is the preferred model? Bootstrapping For the most general model (model 5), compute standard errors for all 3 parameters using the bootstrap method. Reset the seed again to be 100. You should be sampling from the 20 subjects and including the entire 100 rounds for each subject that is chosen. In order to make it run reasonably fast, use B=10.