Extension of the ridge regression technique to non-linear models with additive errors

An extension of the ridge regression technique is proposed for non-linear models, when ill-conditioned data lead to imprecise non-linear least squares estimates. A large sample Bayesian interpretation of the approach is also provided.