What is regularization strength and how does it affect model generalization?

Regularization strength controls how much penalty is applied to model complexity.

Regularization adds constraints to model optimization. Strong regularization (high alpha or lambda) reduces variance but increases bias, potentially underfitting. Weak regularization allows more complex models but risks overfitting. The goal is to find optimal balance using validation techniques.