Evaluation of Neural Network Robust Reliability Using Information-Gap Theory

A novel technique for the evaluation of neural network robustness against uncertainty using a nonprobabilistic approach is presented. Conventional optimization techniques were employed to train multilayer perceptron (MLP) networks, which were then probed with an uncertainty analysis using an information-gap model to quantify the network response to uncertainty in the input data. It is demonstrated that the best performing network on data with low uncertainty is not in general the optimal network on data with a higher degree of input uncertainty. Using the concepts of information-gap theory, this paper develops a theoretical framework for information-gap uncertainty applied to neural networks, and explores the practical application of the procedure to three sample cases. The first consists of a simple two-dimensional (2-D) classification network operating on a known Gaussian distribution, the second a nine-lass vibration classification problem from an aircraft wing, and the third a two-class example from a database of breast cancer incidence

[1]  David J. C. MacKay,et al.  The Evidence Framework Applied to Classification Networks , 1992, Neural Computation.

[2]  Keith Worden,et al.  A novel information-gap technique to assess reliability of neural network-based damage detection , 2006 .

[3]  Ian T. Nabney,et al.  Netlab: Algorithms for Pattern Recognition , 2002 .

[4]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[5]  O. Mangasarian,et al.  Robust linear programming discrimination of two linearly inseparable sets , 1992 .

[6]  David Lowe,et al.  Point-Wise Confidence Interval Estimation by Neural Networks: A Comparative Study based on Automotive Engine Calibration , 1999, Neural Computing & Applications.

[7]  W. A. Wright,et al.  Bayesian approach to neural-network modeling with input uncertainty , 1999, IEEE Trans. Neural Networks.

[8]  Keith Worden,et al.  Experimental validation of a structural health monitoring methodology: Part III. Damage location on an aircraft wing , 2003 .

[9]  A. J. Morris,et al.  Confidence bounds for neural network representations , 1997 .

[10]  S. Hyakin,et al.  Neural Networks: A Comprehensive Foundation , 1994 .

[11]  Lesley F. Wright,et al.  Information Gap Decision Theory: Decisions under Severe Uncertainty , 2004 .

[12]  Tamás D. Gedeon,et al.  Confidence bounds of petrophysical predictions from conventional neural networks , 2002, IEEE Trans. Geosci. Remote. Sens..

[13]  Lionel Tarassenko,et al.  Guide to Neural Computing Applications , 1998 .

[14]  Alan F. Murray,et al.  Confidence estimation methods for neural networks : a practical comparison , 2001, ESANN.

[15]  Heekuck Oh,et al.  Neural Networks for Pattern Recognition , 1993, Adv. Comput..

[16]  Keith Worden,et al.  Information-gap robustness of a neural network regression model , 2004 .

[17]  Luc Jaulin,et al.  Applied Interval Analysis , 2001, Springer London.

[18]  George Chryssolouris,et al.  Confidence interval prediction for neural network models , 1996, IEEE Trans. Neural Networks.

[19]  O. Mangasarian,et al.  Multisurface method of pattern separation for medical diagnosis applied to breast cytology. , 1990, Proceedings of the National Academy of Sciences of the United States of America.

[20]  Christopher M. Bishop,et al.  Neural networks for pattern recognition , 1995 .

[21]  Wlodzislaw Duch,et al.  Uncertainty of data, fuzzy membership functions, and multilayer perceptrons , 2005, IEEE Transactions on Neural Networks.

[22]  David J. C. MacKay,et al.  Information Theory, Inference, and Learning Algorithms , 2004, IEEE Transactions on Information Theory.

[23]  Peter L. Bartlett,et al.  The Sample Complexity of Pattern Classification with Neural Networks: The Size of the Weights is More Important than the Size of the Network , 1998, IEEE Trans. Inf. Theory.

[24]  Keith Worden,et al.  Nonlinearity in Structural Dynamics , 2019 .