Subject: Marketing Analytics II - Targeting and Logistical Regression (Part II)

Q1.

After estimating a logistic regression model using the training data, we plug in the estimated parameter values B, and X values exp (BX) A, and this calculates in the training data: 1+emp(,6X) from the training data, into the question P : O The probability of Y=1 O The probability of Y=0 When comparing the in-sample ts of two logistic regression models estimated using the same data, and the two models have exactly the same number of X variables, we can compare the in-likelihood values and choose: 0 The model with the higher ln-likelihood value 0 The model with the lower ln-likelihood value When estimating a logistic regression model using unbalanced data, where the percentage of positive cases (Y=1) is only 1%, we need to balance the data by: O a) Duplicating the positive cases 0 b) Randomly removing some of the negative cases 0 Either a or b would work 0 Neither 3 nor b would work