Logistic regression and elastic net regression models were used to predict click-through rates for display advertising. Logistic regression with a logit link function and an interaction between variables I5 and I11 performed best with a logloss of 0.4997 and AUC of 0.7371. For elastic net regression, α=0.25 and λ=0.001 achieved the lowest logloss of 0.5131 and an AUC of 0.7199. Median imputation handled missing values better than mean imputation.