Day 15. New project. No model yet. Loan approval. A lender gets a file and has to say yes or no. Approve the wrong person, the bank loses money. Reject the right person, someone loses capital. That is not regression. We are not predicting how much. It is binary classification. Two classes. One target. Target: Loan_Status Y = approve N = reject What the model will see Gender, married, dependents, education, self-employed Applicant income and coapplicant income Loan amount and term Credit history Property area What it will not see Loan_ID. That names the row. It does not describe the person. Expected output A class, and a probability if the algorithm allows it. Not a live credit decision. A first-pass screen on a public table. Same shape as the diabetes project. Different domain. Problem first. Model later. #MachineLearning #Classification #BigBrains