The 15-Day AI/ML Interview Bootcamp
Day 4 — Logistic regression, classification metrics, and the threshold nobody tunes
Logistic regression from scratch: the sigmoid, cross-entropy loss, then confusion matrix, accuracy, precision, recall, F1, ROC and AUC — and why 99% accuracy can be a failing model. With the classification-metrics questions every machine learning interview asks.
You're on call for a payments team. The fraud model shipped last quarter, the dashboard says 99.0% accuracy, and the number has been green every single day. This morning finance forwards the chargeback report: fraud losses are up, not down. You pull the model's decisions for the month and find it flagged eleven payments. Eleven — out of two hundred thousand. It's been right almost every time because almost every payment is legit, and it has been quietly waving the fraud through with the rest. The model isn't broken. The number you've been watching is the wrong number, and the dial that could h…
In this lesson
Unlock the rest of The 15-Day AI/ML Interview Bootcamp
Day 4 — Logistic regression, classification metrics, and the threshold nobody tunes is part of the full course. One payment unlocks every lesson, every other course, and every premium article — for life.