Logistic Regression: Predicting Outcomes with Probability

Final Project: Building and Evaluating a Predictive Logistic Regression Model

In this final assignment, students will select a real-world dataset to predict a binary outcome. The project involves exploratory data analysis, checking for logistic regression assumptions such as linearity of the logit and lack of multicollinearity, and fitting the model using a programming language of choice. Students must interpret the resulting odds ratios, calculate the p-values for predictors, and assess model fit using the Likelihood Ratio test. Performance evaluation must include a confusion matrix, sensitivity, specificity, and an ROC-AUC analysis.