Q1. Attribute selection can help in the following ways: (pick all correct answers)

- Make the data more interpretable

- Enable the design of smaller models

- Help with computational efficiency

- Reduce Overfitting

Q2. Suppose we want to find the 'best-k' attributes in a dataset X, defined as the k-tuple of attributes that give the best possible accuracy among all other k-tuples of attributes, assuming we had to use only k attributes. Then, the following are true: (pick all true statements)

.- Group of answer choices

- Greedy attribute selection can find those k attributes

- It is possible to find the 'best-k' attributes, but it is computationally very expensive

- It is computationally impossible

- Random forest can find the 'best-k' attributes