Regular term can also be added to logistic regression to avoid overfitting.

Question

Options

• ATRUE
• BFALSE

• A
  80% (20)
• B
  20% (5)

Explanation

Regularization terms - such as L1 (Lasso) and L2 (Ridge) penalties - can absolutely be applied to logistic regression, just as they are to linear regression. These terms add a penalty to the loss function that discourages large coefficient values, which constrains model complexity and reduces overfitting. Option B is incorrect because it implies logistic regression is immune to overfitting or that regularization is exclusive to other model types, neither of which is true - logistic regression is a parametric model that can overfit, especially with high-dimensional or sparse data.

Memory tip: Think of regularization as a budget constraint on your model's coefficients - it works the same way regardless of whether the output is continuous (linear regression) or a probability (logistic regression).

No community discussion yet for this question.