Parameters and parametric models

What does the max_iter parameter in logistic regression specify?

• The number of features in the model

• The maximum number of iterations the optimization algorithm can run

• The number of coefficients in the model

• The minimum convergence threshold

Which of the following describes the purpose of the solver parameter in logistic regression?

• It determines the convergence threshold for training

• It specifies the optimization algorithm used to find model coefficients

• It decides the size of the dataset to process

• It sets the data normalization strategy

What does the multi_class='auto' parameter do in logistic regression?

• Forces binary classification regardless of the dataset

• Applies one-vs-rest classification for all solvers

• Chooses the best strategy for handling multiple classes based on the solver

• Converts all target variables to numeric values

What is an example of sparse data? (Select all that apply)

• A dataset where most of the values are zeros or missing

• A dataset with evenly distributed numerical data

• A binary image representation where most pixels are black

• A dataset with dense matrix representation

Which of the following is true about L1 regularization (Lasso)? (Select all that apply)

• It shrinks coefficients closer to zero but never exactly zero

• It adds a penalty proportional to the absolute values of the model coefficients

• It forces some coefficients to become exactly zero

• It uses the sum of squared coefficients as a penalty

When should L1 regularization (Lasso) be preferred over L2 regularization (Ridge)?

• When feature selection is required

• When all predictors are important

• When preventing overfitting is not necessary

• When we want to use all features equally

Why is regularization important in machine learning? (Choose 2)

• It allows the model to learn 'noise' from the data

• It makes the model simpler and easier to interpret

• It prevents overfitting

• It ensures the training data is perfectly memorized

What does a sparse model mean?

• The dataset contains mostly missing or zero values

• The model's coefficients are mostly zero, focusing on a small subset of features

• The model uses all features but regularizes them equally

• The dataset is small and requires minimal computation

Which regularization technique promotes sparsity in the model?

• Ridge regression (L2)

• Lasso regression (L1)

• Elastic Net regression

• None of the above

What are some advantages of sparse models? (Select all that apply)

• They are easy to interpret

• They rely on all features equally

• They handle large datasets efficiently

• They prevent overfitting by ignoring irrelevant features