Interview Practice Questions

What happens when duplicating minority class for data balancing?

How does Random Forest work?

Why are trees in Random Forest made uncorrelated to reduce variance?

Features of classical ML models like logistic regression?

How does dropout work to improve neural network generalization?

How does learning rate tuning work in gradient descent?

How to work with language models in NLP tasks?

How do One-Hot vectors work in feature encoding?

When can logarithm give value 0?

How do 2D optimization landscapes work in ML?

How does fixed step size work in optimization?

What distance metrics are used in machine learning?

How to collect and use user feedback in ML systems?

How to monitor spam classification in real-time chats?

How to use lightweight embeddings for real-time services?

How to choose lightweight embeddings for baseline models?

What types of lightweight embeddings exist?

How to create features based on user message sending speed?

How to use Isolation Forest for anomaly detection?

How to evaluate and improve ML model quality?