Top 50 model training Interview Questions

Implement a CART decision tree from scratch, including split selection, stopping rules, prediction, and evaluation.

Feature Engineeringmodel trainingSupervised LearningAgile DefenseState StreetPublicis Groupe

Explain practical ways to train and evaluate a classifier when the target classes are highly imbalanced.

model trainingSupervised LearningClass ImbalanceIndeedVerisk AnalyticsAtos

Handle rare positive labels in ad fraud detection with the right sampling, loss design, validation, and thresholding strategy.

Feature Engineeringmodel trainingClass ImbalanceNextrollAmerican Family InsuranceSift

Handle severe class imbalance in rare failure prediction while balancing recall, precision, and operational alert volume.

Feature Engineeringmodel trainingClass ImbalanceRivianBaker HughesAvathon

Explain how to train and evaluate a rare event classifier when positives are extremely scarce and false negatives are costly.

model trainingSupervised LearningClass ImbalanceAbnormal AIFortinetSentinelOne

Compare Cross-Entropy and MSE mathematically, then explain how each changes gradient behavior during model training.

loss functionsmodel trainingGradient DescentBMW GroupNetflixTikTok Shop