Iba's home page

\log \cosh(h)の前のMが脱落

Mの定義 -- full model の変数の個数

誤

l_{MF}(\sigma^2,h)= -F_{MF} - \log \cosh(h) - \frac{N}{2} \log {2\sigma^2}

正

l_{MF}(\sigma^2,h)= -F_{MF} - M \log \cosh(h) - \frac{N}{2} \log {2\sigma^2}

{m_i^*}に適当な初期値を与えて、
{m_i^*}から{a_i^*}を求め、
{a_i^*}から{m_i^*}を求め、
収束するまで繰り返す

とありますが、この説明は間違いです。

実際には、(17)式でm_iを 1回 update したあとで、
(18)を解くというのを繰り返しています。方法の狙いや
最適化する式は同じですが、local optimaの様子などには
大きな違いが生じる可能性があるので訂正しておきます。

Belsley,D.A., Kuh,E. and Welsch,R.E. (1980)
Regression Diagnostics,
Wiley, New York.

Geiger,D. and Girosi,F.(1991)
Parallel and deterministic algorithms for MRF's : surface reconstruction and integration,
IEEE Transactions on Pattern Analysis and Machine Intelligence 13 401-412.

George,E.I. and McCulloch,R.E. (1993)
Variable selection via Gibbs sampling,
Journal of the American Statistical Association 88 881-889.

Hopfield,J.J. and Tank,D.W.(1985)
``Neural'' computation of decisions in optimization problems,
Biological Cybernetics 52 141-152.

伊庭幸人(1991)
メトロポリス的モンテカルロ法の擬ベイズ法への応用
-- 変化点問題を例として -- ,
統計数理 39 No.2 225-244.

石黒真木夫 (1985)
ベイズ型重回帰モデル,
統計数理 33 No.1,
特集 -- エントロピー最大化原理の展開と統計モデル -- , pp.8-11.

Ishikawa,M. (1996)
Structural learning with forgetting,
Neural Networks, Vol.9, 509-521.

Kay.J.W. and Titterington D.M. (1986)
Image labelling and statistical analysis of incomplete data Proc. 2nd. Int. Conf. Image processing and Applications Conf. Publ. No.265
London Inst. Elec. Engrs. 44-48.

Kuo,L and Mallick,B. (1996)
Variable selection for regression models,
preprint.

MacKay, D.J.C. (1992)
A Practical Bayesian Framework for Backprop Networks
Neural Computation 4 448-472.

Mitchell,J. and Beauchamp,J.J. (1988)
Bayesian variable selection in linear regression (with discussion),
Journal of the American Statistical Association 83, 1023-1036.

Nowlan,S.J. and Hinton,G.E. (1992)
Simplifying neural networks by soft weight-sharing,
Neural Comutation Vol.4 no.4 pp.473-493.

Opper,M. and Winther,O.
A mean field approach to Bayes learning in feed-foward neural networks
preprint.

Peterson,C. and Hartman,E. (1989)
Explorations of the mean field thoery learning algorithm,
Neural Networks 2 475-494.

Phillips,D.B. and Smith,A.F.M. (1996)
Bayesian model comparison via jump diffusions,
Chap.13 in Markov Chain Monte Carlo in Practice, Eds. Gilks,W.R., Richardson,S. and Spiegelhalter,D.J.,
Chapman and Hall.

Sakai,H. (1993)
An application of BIC type method to harmonic analysis and a new criterion for order determination of an AR process,
ISM Cooperative Research Report 20.

下平英寿 (1993)
モデルの信頼集合と地図によるモデル探索,
統計数理, 41, NO.2, 131-147.

Weigend,A.S. and Rumelhart,D.E. (1991)
Generalization through minimal networks with application to forecasting,
in Computing Science and Statistics:
Proc. of the 23rd symposium on the interface(Ed. E.M.Keramides)
pp.156-163
Interface Foundation,
Fairfax Station,
Va.

Williams,P.M.(1995)
Bayesian regularization and pruning using a Laplace prior,
Neural Computation, Vol.7, No.1 117-143.