Question Bank
Machine Learning interview questions
Naming a model or a metric is only the start of an interview answer. Explain why it fits the task, and connect training, regularization, validation and evaluation to a concrete example. Start with your own explanation, then state its assumptions and limitations.
Showing 20 of 746 questions
When should you use L2 regularization instead of L1 regularization?
Explain how k-fold cross-validation works and what its advantages are.
What is model ensembling and how does it help improve performance?
What is the trade-off between bias and variance in a machine learning model?
What are the main differences between boosting and bagging algorithms?
Explain how Batch Normalization works and what its advantages are.
What is PCA and how is it used in the dimensionality reduction process?
When and why should you use dropout as a regularization method?
Explain what L2 regularization is and how it helps in training models.
Explain how the Adam algorithm works and what its key features are.
Compare L1 and L2 regularization methods. In what cases is each preferred?
Explain the bias-variance tradeoff. How does it affect model selection?
How does momentum gradient descent work and what advantages does it provide?
What is UMAP and how does it differ from t-SNE in the context of dimensionality reduction?
What are the main differences between normalization and standardization of data? When should each approach be used?
How can underfitting and overfitting of a model be diagnosed? What metrics and approaches would you use?
Explain what gradient descent is and how it works.
What is regularization and how does it help combat overfitting?
Compare the ReLU and LeakyReLU activation functions. When should each be used?
Explain what F1 score is and how it differs from accuracy.
Prepare for your next interview with Vibe Interview.
Download Vibe Interview