Interview Practice Questions

How is recall calculated through actual positives?

What are API metrics for independent events?

What types of ML models have you worked with?

What is learning rate and how does it affect training?

Can gradient boosting overfit?

How does Random Forest reduce model variance?

How does CatBoost work with bias and variance?

What role do base trees play in gradient boosting?

When to use cross-validation with limited computational resources?

Why is it important that models in ensemble are uncorrelated?

What types of customer data are collected for ML models?

How to use geographical and behavioral data in feature engineering?

How to extract relevant features from user interaction history?

How to select features that correlate with target variable?

How to identify most reliable and informative features?

How to use lag features to account for previous session influence?

How to incorporate external traffic and congestion data into ML models?

How to collect and use feedback for improving ML models?

What methods exist for handling missing values?

How to use SMOTE for synthetic class generation?