Penalized likelihood regression: General formulation and efficient approximation

By Chong Gu, Young-Ju Kim. 2002

90 citations recorded by the source corpus.

Subfields: Demography, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/442c5c5062ea91beffc99121c4e867d372e83ead

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Demography

In Statistics and Probability

Citations in the corpus

Stable and Efficient Multiple Smoothing Parameter Estimation for Generalized Additive Models (2004, background): “…with good coverage probabilities (see, e.g., Wahba 1990; Gu 2002), but this appealing generality comes at high computational cost—the methods generally require the estimation of one parameter per datum, so the computational cost scales as the cube of the number of data (but see Gu and Kim 2002).”
Low‐Rank Scale‐Invariant Tensor Product Smooths for Generalized Additive Mixed Models (2006, unknown): “In the case of some spline-based models Gu and Kim (2002) derive the result that basis size should be kn2/9, where n is sample size, but unfortunately k is an unknown constant, which really depends on the unknown truth being estimated.”
Smoothing spline Gaussian regression: more scalable computation via efficient approximation (2004, methodology): “Algorithms for the computation of ηλ were developed by Gu et al. (1989) and Gu and Wahba (1991) and employed a certain numerical structure that is not shared by the approximation in Hq=NJ ⊕ span{RJ.”
A simple approach for varying-coefficient model selection (2009, background): “Quantification of the coefficients can be done in a hypothesis testing framework, see Huang et al. (2002) and Fan and Huang (2006).”
ON CONFIDENCE INTERVALS FOR GENERALIZED ADDITIVE MODELS BASED ON PENALIZED REGRESSION SPLINES (2006, background): “In this context recent work by Gu & Kim (2002) on the rate at which the parameter space dimension should increase with sample size for certain spline models may be helpful.”