434. L1 Sparsity and Bias-Variance
medium
L1 regularization is applied to a regression model with 100 features. After training, 85 features have zero coefficients. How does this affect the bias-variance tradeoff compared to the unregularized model?
L1 regularization is applied to a regression model with 100 features. After training, 85 features have zero coefficients. How does this affect the bias-variance tradeoff compared to the unregularized model?