Patent
1997-01-08
1998-07-07
Davis, George B.
395 20, 395 21, 395 23, G06F 1518
Patent
active
057781532
ABSTRACT:
A neural network, which may be implemented either in hardware or software, is constructed of neurons or neuron circuits each having only one significant processing element in the form of an adder. Each neural network further includes circuits for applying a logarithmic function to its inputs and for applying an inverse-logarithmic function to the outputs of its neurons. The neural network utilizes a training algorithm which does not require repetitive training and which yields a global minimum to each given set of input vectors.
REFERENCES:
patent: 3551663 (1970-12-01), Herron et al.
patent: 3748451 (1973-07-01), Ingwersen
patent: 3922536 (1975-11-01), Hampel et al.
patent: 3941990 (1976-03-01), Rabasse
patent: 3967100 (1976-06-01), Shimomura
patent: 4156922 (1979-05-01), Majerski et al.
patent: 4626825 (1986-12-01), Burleson et al.
patent: 4849925 (1989-07-01), Peckerar et al.
patent: 4905143 (1990-02-01), Takahashi et al.
patent: 4949292 (1990-08-01), Hoshino et al.
patent: 5042001 (1991-08-01), Brightman et al.
patent: 5068816 (1991-11-01), Noetzel
patent: 5072130 (1991-12-01), Dobson
patent: 5073867 (1991-12-01), Murphy et al.
patent: 5109524 (1992-04-01), Wagner et al.
patent: 5113367 (1992-05-01), Marrian et al.
patent: 5220559 (1993-06-01), Tsuzuki et al.
patent: 5278945 (1994-01-01), Basehore et al.
patent: 5331582 (1994-07-01), Sudo et al.
patent: 5337266 (1994-08-01), Arnold
patent: 5343254 (1994-08-01), Wada et al.
patent: 5359551 (1994-10-01), Pickett
patent: 5365465 (1994-11-01), Larson
patent: 5381513 (1995-01-01), Tsuboka
patent: 5444819 (1995-08-01), Negishi
patent: 5450522 (1995-09-01), Hermansky et al.
patent: 5517667 (1996-05-01), Wang
patent: 5553012 (1996-09-01), Buss et al.
Algorithm Design for a 30 bit Integrated Logarithmic Processor by David M. Lewis and Lawrence K. Yu, Department of Electrical Engineering, University of Toronto, Proceeding 9th Symposium on Computer Arithmetic, 1989, IEEE Comp. Soc. Press, pp. 192-199.
An Architecture for Addition and Subtraction of Long Word Length Numbers in the Logarithmic Number System by David M. Lewis, Member, IEEE, IEEE Transactions on Computers, vol. 39, No. 11. Nov. 1990, pp. 1325-1336.
A 30-b Integrated Logarithmic Number System Processor by Lawrence K. Yu, Member, IEEE, and David M. Lewis, Member, IEEE, IEEE Journal of Solid-State Circuits, vol. 26, No. 10, Oct. 1991, pp. 1433-1440.
An Accurate LNS Arithmetic Unit Using Interleaved Memory Function Interpolator by David M. Lewis, Department of Electrical Engineering, University of Toronto, Proceeding 11th Symposium on Computer Arithmetic. 1993, IEEE Comp. Soc. Press, pp. 2-9.
Interleaved Memory Function Interpolators with Application to an Acuurate LNS Arithmetic Unit by David M. Lewis, Member, IEEE, IEEE Transactions on Computers, vol. 43, No. 8, Aug. 1994, pp. 974-982.
Table-Lookup Algorithms for Elementary Functions and Their Error Analysis by Ping Tak Peter Tang, Matematics and Computer Schience Division, Argonne National Laboratory, 9700 S. Cass Ave., Argonne, IL 60439-4801, Proceeding 10th Symposium on Computer Arithmetic, Jun. 1991, pp. 232-236.
Applying Features of IEEE 754 to Sign/Logarithm Arithmetic by Mark G. Arnold, Member, IEEE, Thomas A. Bailey, Member, IEEE, John R. Cowles, and Mark D. Winkel, IEEE Transactions on Computers, vol. 41, No. 8, Aug. 1992, pp. 1040-1050.
D8.13 Improved Accuracy for Logarithmic Addition in DSP Applications by Mark G. Arnold, John Cowles, and Thomas Bailey, Computer Science Department, University of Wyoming, Laramie, WY, ICASSP 88: Int. Conf. on Acoustics, Speech and Signal Processing, vol. 3 pp. 1714-1717.
Comments on "An Architecture for Addition and Subtraction of Long Word Length Numbers in the Logarithmic Number System".sup.1 by M. Arnold, T. Bailey and J. Cowles, IEEE Transactions on Computers, vol. 41, No. 6, Jun. 1992, pp. 786-788.
Redundant Logarithmic Arithmetic, Mark G. Arnold, Member IEEE, Thomas A. Bailey, Member IEEE, John R. Cowles, and Jerry J. Cupal, Members IEEE, IEEE Transactions on Computers, vol. 39, No. 8, Aug. 1990, pp. 1077-1086.
The Efficient Implimentation and Analysis of a Hybrid Number System Processor, Fang-shi Lai, IEEE Transactions on Circuits and Systems, II: Analog and Digital Signal Processing. vol. 40, No. 6, Jun. 1993, pp. 382-392.
Polynomial and Standard Higher Order Neural Network, Chir-Ho Chang, Jin-Ling Lin, and J.Y. Cheung, Electrical Engineering and Computer Science, University of Oklahoma, 1993 IEEE International Conference on Neural Networks, Mar. 28-Apr. 1, 1993, pp. 989-994.
A Digital Neuron-Type Processor and Its VLSI Design, Mahmoud K. Habib, Member, IEEE, and H. Akel, 8090 IEEE Transactions on Circuits and Systems 36(1989) May. No. 5, New York, US, pp. 739-746.
A Neural Feed-Forward Network with a Polynomial Nonlinearity, Nils Hoffmann, Electronics Institute, Building 349, Technical University of Denmark, DK-2800 Lyngby, Denmark, Neural Networks for Signal Processing, Proceedings of the IEEE-SP Workshop, Aug. 31-Sep. 2, 199?, pp. 49-58.
A Polynomial time Algorithm for Generating Neural Networkes for Classification Problems, Asim Roy and Somnath Mukhopadhyay, Dept. of Decision and Information Systems, Arizona State University, IJCNN, 1992, 0-7803-0559-0/92 IEEE, pp. I-147-I-152.
Polynomial Functions Can Be Realized by Finite Size Multilayer Feedforward Neural Networks, Naohiro TODA, Ken-ichi Funahashi and Shiro Usui, Department of Information and Computer Sciences, Toyohashi University of Technology, Tempaku, Toyohashi 441, Japan, 1991 IEEE International Joint Conference on Neural Networks, vol. 1of3,The Wetsin Stamford and Westin Plaza, 18-21, Nov. 1991, Singapore.
An Artificial Neural Networks for Approximating Polynomial Functions, Behnam Malakooti and YingQing Zhou, Department of Ssytems Engineering, Center for Automation and Intelligent Systmes Research, Case Western Reserve University, Cleveland, Ohio,International Joint Conference on Neural Networks, Jun. 1992, pp. III-966-III-971.
Robot Kinematics Learning Computations Using Polynomial Neural Networks, C.L. Philip Chen and A.D. McAulay, Department of Computer Science and Engineering, Wright State University, Dayton, OH 45435, Proceedings of the 1991 IEEE, International Cof. on Robotics and Automation, Sacramento, CA, Apr. 1991, pp. 2638-2643.
Backpropagation Based on the Logarithmic Error Function and Elimination of Local Minima, Kiyotoshi Matsuoka and Jianqiang Yi, Department of Control Engineering, Kyushu Institute of Technology, Sensul 1-1. Tobata. Kitakyushu, 804 Japan, CH3065-0/91/0000-1117 IEEE, pp. 1117-1122.
Output Weight Optimization for the Multi-Layer Perceptron by M.T. Manry, Xiujun Guan, S.J. Apollo, L.S. Allen, W.D. Lyle, and W. Gong, Department of Electrical Engineering, University of Texas at Arlington, Arlington, Texas, Conference Record of the Twenty Sixth Asilomar conf. on Signals, Systems and Computers, Oct. 26-28, 1992, pp. 502-506.
Generalization and Learning in Volterra and Radial Basis Function Networks: A Theoretical Analysis, Sean B. Holden and Peter J. W. Rayner, Cambridge University Engineering Department, Trumpington Street, Cambridge CB2 1PZ, U.K.A, 0-7803-0532-9/92, 1992 IEEE, pp. II-273-II276.
On the Design Principlpes of the Functional Link Nets, Jun Wang and Edilberto P. Teixeira, Department of Systems Engineering, Case Western Reserve University, Cleveland, Ohio 44106, pp. 613-616, Ieee International Conf. on Systems Engineering, Aug. 9-11, 1990, Vista International Hotel, Pittsburgh, Pennsylvania, IEEE Catalog No:90CH2872-0.
A Neural Network Systems Component, Dean Mueller and Dan Hammerstrom, Adaptive Solutions, Inc, 1400 NW Compton Drive Suite 340, Beaverton, Oregon 97006,pp. 1258-1264, Neural Networks, 1993 International Conference.
Parallel Coprocessor for Kohonen's Self-Organizing Neural Network, Jukka Saarinen, Martti Lindroos, Jouni Tomberg and Kimmo Kaski, Tampere University of Technology, Microelectronics Laboratory, P.O. Box 527, SF-33101 Tampere, Finland, pp. 537-542 Parallel Processing, 1992 Symposium.
Hoffmann et al., "A Neural Feed-Forward Network with a Polynom
Davis George B.
Gauger James E.
Motorola Inc.
LandOfFree
Neural network utilizing logarithmic function and method of usin does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Neural network utilizing logarithmic function and method of usin, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Neural network utilizing logarithmic function and method of usin will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-1215920