You are a data analyst at a bank and are tasked with creating a classification algorithm to predict which customers will accept a personal loan offer. You have partitioned the data and used the training data to fit coefficients of a logistic regression model using the predictor variables experience (i.e. years at the bank), income (in $1000s), family size, average credit card balance (CCAvg), and whether or not the individual accepted a personal loan offer (0: non-acceptor, 1: acceptor).

You have used the fitted regression equation to calculate the likelihood score (P, in column G) from the logistic regression equation that is being used to predict personal loan acceptance. In this case, you will evaluate the performance of the classifier and determine the best cutoff threshold to use when predicting loan acceptance.

 

Transcribed Image Text:

Coefficients b0 Experience Income Family CCAvg Education -13.92 0.010488 0.055902 0.738787 0.118769 1.64886 Cutoff Threshold T 0.5 Predicted Value 1 Actual 1113 26 Value 50 61 Coefficients b0 Experience Income Family CCAvg Education -13.92 0.010488 0.055902 0.738787 0.118769 1.64886 Cutoff Threshold T 0.5 Predicted Value 1 Actual 1113 26 Value 50 61

Answer rating: 100% (QA)

PREDICTED negative positive ACTUAL negative TN FP positive FN TP 1 The overall accuracy of the class... View the full answer