Experimental design and decision support

Publisher Summary This chapter explores, designs, develops, and implements fuzzy neural network models for organizing and learning experimental data and knowledge that can be fruitful in producing fuzzy neural experimental decision support systems for various domains. It presents a new concept mandatory for experimental design and decision. The analysis of the results is difficult through statistical techniques, and needs restrictive assumptions for the problem to be analytically formulated, or a prohibitively large computational endeavor. These shortcomings lead to the application of a combination of other methods or tools such as genetic algorithms, neural networks, fuzzy logic, and machine learning. This chapter discusses a number of concepts that include surface approximation, sampling, universal approximation, and fuzzy neural networks. Localized sampling and adaptive reconstruction are the two ideas conceived from these concepts. These two ideas are the main learning engine and allow an experimental design and decision support model to develop and reconstruct by learning novel features from the localized input and output sequence samples. These proposed concepts provide a nonlinear, model-free, multivariate surface data fitting learning algorithm, which automatically determines the functional relationship between inputs and outputs directly from data without a hypothesized functional form. Both neural networks and fuzzy systems offer the property of universal approximation. They are, therefore, perfect for solving and representing experimental design problems that provide a means to decision support. The relation between neural networks and fuzzy systems provides a link that connects two seemingly different research areas.

[1]  Robert M. Farber,et al.  How Neural Nets Work , 1987, NIPS.

[2]  M. J. D. Powell,et al.  Radial basis functions for multivariable interpolation: a review , 1987 .

[3]  Donald F. Specht,et al.  A general regression neural network , 1991, IEEE Trans. Neural Networks.

[4]  Pierre Baldi,et al.  Computing with Arrays of Bell-Shaped and Sigmoid Functions , 1990, NIPS.

[5]  R. Fisher,et al.  Statistical Methods for Research Workers , 1930, Nature.

[6]  D. F. Morrison,et al.  Multivariate Statistical Methods , 1968 .

[7]  Madhan Shridhar Phadke,et al.  Quality Engineering Using Robust Design , 1989 .

[8]  Neil R. Ullman,et al.  Signal-to-noise ratios, performance criteria, and transformations , 1988 .

[9]  P. A. Jokinen.  On the relations between radial basis function networks and fuzzy systems , 1992, [Proceedings 1992] IJCNN International Joint Conference on Neural Networks.

[10]  Thomas M. Cover,et al.  Geometrical and Statistical Properties of Systems of Linear Inequalities with Applications in Pattern Recognition , 1965, IEEE Trans. Electron. Comput..

[11]  Teuvo Kohonen,et al.  Representation of Associated Data by Matrix Operators , 1973, IEEE Transactions on Computers.

[12]  Genichi Taguchi,et al.  Quality Engineering through Design Optimization , 1989 .

[13]  J. Pignatiello,et al.  Discussion: Off-Line Quality Control, Parameter Design, and the Taguchi Method , 1985 .

[14]  A. Kaufmann,et al.  Introduction to fuzzy arithmetic : theory and applications , 1986 .

[15]  Daryl Pregibon,et al.  A data analysis strategy for quality engineering experiments , 1986, AT&T Technical Journal.

[16]  George Cybenko,et al.  Approximation by superpositions of a sigmoidal function , 1992, Math. Control. Signals Syst..

[17]  Stephen Grossberg,et al.  Nonlinear neural networks: Principles, mechanisms, and architectures , 1988, Neural Networks.

[18]  Raghu N. Kacker,et al.  Graphical and computer-aided approaches to plan experiments , 1989 .

[19]  Henry P. Wynn,et al.  Quality through design : experimental design, off-line quality control and Taguchi's contributions , 1991 .

[20]  R. N. Kackar.  Off-Line Quality Control, Parameter Design, and the Taguchi Method , 1985 .

[21]  D. Broomhead,et al.  Radial Basis Functions, Multi-Variable Functional Interpolation and Adaptive Networks , 1988 .

[22]  Chris Bishop,et al.  Improving the Generalization Properties of Radial Basis Function Neural Networks , 1991, Neural Computation.

[23]  J. Senturia.  System of Experimental Design (Vol. 2) , 1989 .

[24]  Stephen Jones,et al.  [Testing in Industrial Experiments with Ordered Categorical Data]: Discussion , 1986 .

[25]  D. Pregibon,et al.  REX: an Expert System for Regression Analysis , 1984 .

[26]  A. C. Shoemaker,et al.  Performance Measures Independent of Adjustment: An Explanation and Extension of Taguchi's Signal-to-Noise Ratios , 1987 .

[27]  Arthur M. Schneiderman,et al.  Optimum Quality Costs and Zero Defects: Are They Contradictory Concepts? , 1986 .

[28]  A. Khuri,et al.  Simultaneous Optimization of Multiple Responses Represented by Polynomial Regression Functions , 1981 .

[29]  André I. Khuri,et al.  A Test for Lack of Fit of a Linear Multiresponse Model , 1985 .

[30]  Takayuki Ito,et al.  Neocognitron: A neural network model for a mechanism of visual pattern recognition , 1983, IEEE Transactions on Systems, Man, and Cybernetics.

[31]  J. Stuart Hunter,et al.  Statistical Design Applied to Product Design , 1985 .

[32]  Visakan Kadirkamanathan,et al.  Sequential Adaptation of Radial Basis Function Neural Networks and its Application to Time-series Prediction , 1990, NIPS 1990.

[33]  R. Plackett,et al.  THE DESIGN OF OPTIMUM MULTIFACTORIAL EXPERIMENTS , 1946 .

[34]  Anne C. Shoemaker,et al.  Robust design: A cost-effective method for improving manufacturing processes , 1986, AT&T Technical Journal.

[35]  A. Sarajedini,et al.  The best of both worlds: Casasent networks integrate multilayer perceptrons and radial basis functions , 1992, [Proceedings 1992] IJCNN International Joint Conference on Neural Networks.

[36]  James M. Lucas,et al.  Exponentially weighted moving average control schemes: Properties and enhancements , 1990 .

[37]  G. Lorentz.  Approximation of Functions , 1966 .

[38]  Madan M. Gupta.  Fuzzy logic and neural networks , 1992, [Proceedings 1992] IEEE International Conference on Systems Engineering.

[39]  John F. Kolen,et al.  Learning in parallel distributed processing networks: Computational complexity and information content , 1991, IEEE Trans. Syst. Man Cybern..

[40]  Lotfi A. Zadeh,et al.  Outline of a New Approach to the Analysis of Complex Systems and Decision Processes , 1973, IEEE Trans. Syst. Man Cybern..

[41]  T. R. Bement,et al.  Taguchi techniques for quality engineering , 1995 .

[42]  G. Alexits.  Approximation theory , 1983 .

[43]  I. Johnstone,et al.  Projection-Based Approximation and a Duality with Kernel Methods , 1989 .

[44]  A. Bendell,et al.  Taguchi methods : applications in world industry , 1989 .

[45]  Jooyoung Park,et al.  Universal Approximation Using Radial-Basis-Function Networks , 1991, Neural Computation.

[46]  Conrad A. Fung,et al.  An explanation and critique of taguchi's contributions to quality engineering , 1988 .

[47]  Jerome Sacks,et al.  Computer Experiments for Quality Control by Parameter Design , 1990 .

[48]  Jean-Pierre Martens,et al.  A fast and robust learning algorithm for feedforward neural networks , 1991, Neural Networks.

[49]  Shang-Liang Chen,et al.  Orthogonal least squares learning algorithm for radial basis function networks , 1991, IEEE Trans. Neural Networks.

[50]  Bart Kosko,et al.  ADAPTIVE INFERENCE IN FUZZY KNOWLEDGE NETWORKS , 1993 .

[51]  Kumpati S. Narendra,et al.  Identification and control of dynamical systems using neural networks , 1990, IEEE Trans. Neural Networks.

[52]  F. Girosi.  Some extensions of radial basis functions and their applications in artificial intelligence , 1992 .

[53]  P. K. Simpson.  Fuzzy Min-Max Neural Networks-Part 1 : Classification , 1992 .

[54]  William Q. Meeker,et al.  A Computer Program for Evaluating and Comparing Experimental Designs and Some Applications , 1975 .

[55]  Soren Bisgaard.  Quality Engineering and Taguchi Methods: A Perspective , 1990 .

[56]  James D. Keeler,et al.  Algorithms for Better Representation and Faster Learning in Radial Basis Function Networks , 1989, NIPS.

[57]  Takeshi Yamakawa,et al.  A fuzzy inference engine in nonlinear analog mode and its application to a fuzzy logic control , 1993, IEEE Trans. Neural Networks.

[58]  James D. Keeler,et al.  Layered Neural Networks with Gaussian Hidden Units as Universal Approximations , 1990, Neural Computation.

[59]  J. S. Hunter,et al.  Statistics for experimenters : an introduction to design, data analysis, and model building , 1979 .

[60]  N. Logothetis,et al.  Off‐line quality control and ill‐designed data , 1987 .

[61]  H. Zimmermann,et al.  Fuzzy Set Theory and Its Applications , 1993 .

[62]  Bart Kosko,et al.  Fuzzy knowledge combination , 1986, Int. J. Intell. Syst..

[63]  Douglas M. Bates,et al.  Multiresponse Estimation With Special Application to Linear Systems of Differential Equations , 1985 .

[64]  John F. MacGregor,et al.  ASQC Chemical Division Technical Conference 1971 Prize Winning Paper Some Problems Associated with the Analysis of Multiresponse Data , 1973 .

[65]  Richard K. Miller,et al.  Automated Inspection and Quality Assurance , 1989 .

[66]  Madan Gupta,et al.  Multivariable Structure of Fuzzy Control Systems , 1986, IEEE Transactions on Systems, Man, and Cybernetics.

[67]  Bart Kosko,et al.  Neural networks for signal processing , 1992 .

[68]  Lyle H. Ungar,et al.  Using radial basis functions to approximate a function and its error bounds , 1992, IEEE Trans. Neural Networks.

[69]  Raghu N. Kacker,et al.  Performance measures independent of adjustment , 1987 .

[70]  Joseph J. Pignatiello,et al.  An experimental design strategy for designing robust systems using discrete-event simulation , 1991, Simul..