The data for this assignment is contained on the spreadsheet titled “Data” in the Excel Workbook titled “HW2 Data.”
statistics
Description
Homework #3
Please use the same data as was used for HW #2. That is, the data for this assignment is
contained on the spreadsheet titled “Data” in the Excel Workbook titled “HW2
Data.” There are 800 observations split
into 4 subsets; namely, Dataset 1, Dataset 2, Dataset3, and Dataset 4 (i.e.,
see Column A).
Let’s create a poor man’s version of model averaging.
Use Dataset1 and Dataset2 combined to fit a full model. Use Dataset3 as the validation set. That is, take the model developed from the
combination of Dataset1 and Datset2 and score it against Dataset3. Find the R-sq for the scored data; let’s call
it 
(This would be model 3). Repeat this process by combining Dataset1 and
Dataset3 as the training set and scoring Dataset2 [
] (This would be model
2); as well as using Dataset2 and Dataset3 and scoring Dataset1 [
] (This would be model
1). Develop the final model by averaging
Models 1,2, and 3. That is, weight Model
1 by 
; Model 2 by 
; and Model 3 by 
.
Report your model averaging result; i.e.,
the resultant final model.
Validate the performance of the resultant final model using
the Test Set (i.e., Dataset4).
