In Using Both the Linear Probability Regression and the Logistic $\hat{y}$

Multiple Choice

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 $\hat{y}$ to binary predictions?

A) There is no relevant outcome changing ? to a binary variable.
B) By comparing $\hat{y}$ to y, the validation results are not able to be validated.
C) By comparing $\hat{y}$ to y, the accuracy of the model is 100%, but can change with a larger validation set.
D) By comparing $\hat{y}$ to y, the accuracy of the model is 100% and will not change based on a larger validation set.