Machine Learning / AI

What is linear regression and how is it trained?

Why can a model with ROC-AUC 0.2 achieve ROC-AUC 0.8?

Calculate ROC-AUC for a constant model and a random model with a class ratio of 9 to 1.

What are the differences between XGBoost, LightGBM, and CatBoost?

How do decision trees work and what happens in the absence of constraints?

What is logistic regression and how does it work?

Explain how the gradient boosting method on decision trees works.

Explain the principle of regularization in linear models and its impact on overfitting.

What happens when the learning rate in gradient descent is too high or too low?

How do you select features in a model? There is boosting and 5000 features.

How does a linear model handle multicollinearity of features?

Describe the feature selection process using decision trees.

Can you explain how TPR and FPR are calculated at each threshold?

What metrics are suitable for evaluating models with imbalanced classes?

Can you explain in more detail how exactly the anti-gradient is calculated and how it is used to build the next tree?

Can you explain in more detail how the numerical value of feature importance is calculated in a decision tree?