手法を比較

	経験ベイズ	混合効果モデル	リッジ回帰
分野	ベイズ	統計学	機械学習
系統	Bayesian methods	Regression model	Machine learning
提唱年	—	1982	1970
提唱者	Herbert Robbins (1956); Bradley Efron & Carl Morris (1973)	Laird & Ware	Hoerl, A.E. & Kennard, R.W.
種類	Empirical Bayes estimator	Mixed effects regression	L2-regularized linear regression
原典	Robbins, H. (1956). An empirical Bayes approach to statistics. In J. Neyman (Ed.), Proceedings of the Third Berkeley Symposium on Mathematical Statistics and Probability, Vol. 1 (pp. 157–164). University of California Press.	Laird, N. M., & Ware, J. H. (1982). Random-effects models for longitudinal data. Biometrics, 38(4), 963–974.	Hoerl, A.E. & Kennard, R.W. (1970). Ridge Regression: Biased Estimation for Nonorthogonal Problems. Technometrics, 12(1), 55–67.
別名	EB, empirical Bayes estimation, marginal likelihood estimation, James-Stein shrinkage	LME, LMM, mixed model, random effects model	Ridge Regresyonu, ridge regresyonu, L2-regularized regression, Tikhonov regularization
関連	4	4	4
概要	Empirical Bayes (EB) is an estimation strategy, introduced by Herbert Robbins in 1956 and developed into practical shrinkage estimators by Bradley Efron and Carl Morris in 1973, in which the hyperparameters of the prior distribution are estimated from the observed data via the marginal likelihood rather than specified in advance. The resulting posterior retains a Bayesian structure but substitutes data-driven hyperparameters for subjective ones, bridging frequentist shrinkage and full Bayesian inference.	A mixed effects model (or linear mixed model) extends ordinary regression by including both fixed effects — population-level parameters shared by all observations — and random effects that capture subject-, group-, or cluster-level variability. It is the standard tool for repeated-measures, longitudinal, and multilevel data where observations within the same unit are correlated.	Ridge Regression is an L2-regularized linear regression method, introduced by Arthur Hoerl and Robert Kennard in 1970, that reduces multicollinearity by adding a penalty on the size of the coefficients. It shrinks coefficients toward zero without setting any of them exactly to zero, producing more stable estimates when predictors are highly correlated.