Bias reduction in skewed binary classification with Bayesian neural networks
作者： P.J.G Lisboa , A Vellido , H Wong
DOI: 10.1016/S0893-6080(00)00022-8
关键词:
摘要: The Bayesian evidence framework has become a standard of good practice for neural network estimation class conditional probabilities. In this approach the probability is marginalised over distribution weights, which usually approximated by an analytical expression that moderates output towards midrange. paper, it shown calibration considerably improved marginalising to prior distribution. Moreover, marginalisation midrange can seriously bias estimates probabilities calculated from framework. This especially case in modelling censored data.
elsevier.com
sci-hub.se 参考文章(8)
James E. Davis, Dealing with Censorship. National Council of Teachers of English, 1111 Kenyon Road, Urbana, Illinois 61801 (Stock No. 10622, $6.50 member, $7.50 non-member). ,(1979)
K. Meehan, A. Vellido, P.J.G. Lisboa, Segmentation of the on-line shopping market using neural networks Expert Systems with Applications. ,vol. 17, pp. 303- 314 ,(1999) , 10.1016/S0957-4174(99)00042-1
Sauchi Stephen Lee, Regularization in skewed binary classification Computational Statistics. ,vol. 14, pp. 277- 292 ,(1999) , 10.1007/S001800050018
David J. C. MacKay, The evidence framework applied to classification networks Neural Computation. ,vol. 4, pp. 720- 736 ,(1992) , 10.1162/NECO.1992.4.5.720
D. Lowe, A.R. Webb, Optimized feature extraction and the Bayes decision in feed-forward classifier networks IEEE Transactions on Pattern Analysis and Machine Intelligence. ,vol. 13, pp. 355- 364 ,(1991) , 10.1109/34.88570
Elia Biganzoli, Patrizia Boracchi, Luigi Mariani, Ettore Marubini, Feed forward neural networks for the analysis of censored survival data: A partial logistic regression approach Statistics in Medicine. ,vol. 17, pp. 1169- 1186 ,(1998) , 10.1002/(SICI)1097-0258(19980530)17:10<1169::AID-SIM796>3.0.CO;2-D
H. Wong, P. Harris, P.J.G. Lisboa, S.P.J. Kirby, R. Swindell, Dealing with censorship in neural network models international joint conference on neural network. ,vol. 5, pp. 3702- 3706 ,(1999) , 10.1109/IJCNN.1999.836273
Lionel Tarassenko, Guide to Neural Computing Applications ,(1998)