Preventing Overfitting\n\nRegularization adds a penalty to large weights in the model to prevent them from becoming too complex. Lasso (L1) and Ridge (L2) are the standards.