Bimodal Distribution Removal

A number of methods for cleaning up noisy training sets to improve generalisation have been proposed recently. Most of these methods perform well on artificially noisy data, but less well on real world data where it is difficult to distinguish between noisy data points from valid but rare data points.

[1]  Jeffrey A. Joines,et al.  Improved generalization using robust cost functions , 1992, [Proceedings 1992] IJCNN International Joint Conference on Neural Networks.

[2]  Halbert White,et al.  Learning in Artificial Neural Networks: A Statistical Perspective , 1989, Neural Computation.

[3]  Elie Bienenstock,et al.  Neural Networks and the Bias/Variance Dilemma , 1992, Neural Computation.

[4]  H. White,et al.  A Unified Theory of Estimation and Inference for Nonlinear Dynamic Models , 1988 .

[5]  T. Ash,et al.  Dynamic node creation in backpropagation networks , 1989, International 1989 Joint Conference on Neural Networks.

[6]  Jocelyn Sietsma,et al.  Creating artificial neural networks that generalize , 1991, Neural Networks.