Question: In using both the linear probability regression and the logistic regression models for n = 40, the following table is the analysis of the holdout
In using both the linear probability regression and the logistic regression models for n = 40, the following table is the analysis of the holdout method. Based on the table, what is the impact of changing the to binary predictions?
| Observation | Y | (Model 1) | (Model 2) | |||||
| 18 | 1 | 1.0263 | 0.9907 | |||||
| 19 | 0 | 0.1987 | 0.3032 | |||||
| : | : | : | : | |||||
| 40 | 1 | 0.1509 | 0.1989 | |||||
| Accuracy | 100 | % | 100 | % | ||||
Multiple Choice
-
There is no relevant outcome changing to a binary variable.
-
By comparing to y, the validation results are not able to be validated.
-
By comparing to y, the accuracy of the model is 100%, but can change with a larger validation set.
-
By comparing to y, the accuracy of the model is 100% and will not change based on a larger validation set.
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
