Hyperparameter Tuning Flashcards

Grid and Random Search Cross Validation (CV) (5 cards)

1
Q

Permutation set

A

A set of elements arranged in a specific order. In machine learning, permutation is important in evaluating the effect of each feature on model performance by shuffling its values.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Polynomial Features

A

Used to convert features to a higher degree.
If there are 2 features [a, b], then degree-2 polynomial features are [1, a, b, a^2, ab, b^2].

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

When should polynomial features be used?

A

Use when the relationship between the target variable and independent variables is nonlinear. This can be determined with a negative R2score

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

T/F: Hypberperameters are learned from the data?

A

False: Hyperperameters are used explicitly to tune the model.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Data leakage can be caused by

A

Treating the training and test data together before splitting.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly