Lasso Regression
Linear regression with L1 regularization and automatic feature selection
Lasso adds an L1 penalty that drives irrelevant feature coefficients exactly to zero, performing implicit feature selection. Useful when you suspect only a subset of features are relevant.
When to use:
- High-dimensional data where most features are expected to be irrelevant
- When a sparse, interpretable model is desired
- Feature selection combined with regression
Input: Tabular data with the feature columns defined during training Output: Continuous predicted value
Model Settings (set during training, used at inference)
Alpha (default: 1.0) Regularization strength. Higher values produce sparser models with fewer non-zero coefficients.
Max Iter (default: 1000) Maximum coordinate descent iterations for convergence.
Fit Intercept (default: true) Whether to include a bias term.
Selection (default: cyclic)
Order of coefficient updates. random can converge faster on some problems.
Inference Settings
No dedicated inference-time settings. Only non-zero coefficients from training contribute to predictions.