Method for training a statistical classifier with reduced tenden

Details Method for training a statistical classifier with reduced tenden Method for training a statistical classifier with reduced tenden

1995-08-08

1999-05-11

Downs, Robert W.

Data processing: artificial intelligence

Neural network

Learning method

706 20, 382155, 382157, 382190, G06F15/18

Patent

active

059038845

ABSTRACT:
To prevent overfitting a neural network to a finite set of training samples, random distortions are dynamically applied to the samples each time they are applied to the network during a training session. A plurality of different types of distortions can be applied, which are randomly selected each time a sample is applied to the network. Alternatively, a combination of two or more types of distortion can be applied each time, with the amount of distortion being randomly varied for each type.

REFERENCES:
patent: 5058180 (1991-10-01), Khan
patent: 5239594 (1993-08-01), Yoda
patent: 5276771 (1994-01-01), Manukian et al.
patent: 5297237 (1994-03-01), Masuoka et al.
patent: 5299284 (1994-03-01), Roy
patent: 5333210 (1994-07-01), Spirkovska et al.
patent: 5359699 (1994-10-01), Tong et al.
patent: 5390285 (1995-02-01), Wood
patent: 5469528 (1995-11-01), Douglas et al.
patent: 5504780 (1996-04-01), Alspector et al.
patent: 5546503 (1996-08-01), Abe et al.
patent: 5572628 (1996-11-01), Denker et al.
patent: 5590218 (1996-12-01), Ornstein
patent: 5640468 (1997-06-01), Hsu
M. Caudill, "Neural Network Training Tips and Techniques," Al Expert, pp. 56-61, Jan. 1991.
D. Whitley and N. Karunanithi, "Generalization in Feed Forward Neural Networks," Int'l. Joint Conf. on Neural Networks, vol. 2, pp. 77-82, Jul. 1991.
P. Koistinen and L. Holmstrom, "Kernel regression and backpropagation training with noise," 1991 IEEE Int'l. Joint Conf. on Neural Networks, vol. 1, pp. 367-372, Nov. 1991.
Z.M.V. Kovacs and R. Guerrieri, "A Generalization Technique for Nearest-Neighbor Classifiers," 1991 IEEE Int'l. Joint Conf. on Neural Networks, vol. 2, pp. 1782-1788, Nov. 1991.
L. Holmstrom and P. Koistinen, "Using Additive Noise in Back-Propagation Training," IEEE Trans. on Neural Networks, vol. 3(1), pp. 24-38, Jan. 1992.
R.D. Clay and C.H. Sequin, "Fault Tolerance Tranining Improves Generalization and Robustness," Int'l. Joint Conf. on Neural Networks, vol. 1, pp. 769-774, Jun. 1992.
R. Aloni-Lavi, et al., "A BP Variant with Improved Convergence and Generalization," Int'l. Joint Conf. on Neural Networks, vol. 1, pp. 932-937, Jun. 1992.
R. Reed, et al., "An Equivalence Between Sigmoidal Gain Scaling and Training with Noisy (jittered) input data," RNNS/IEEE Symp. on Neuroinformatics and Neurocomputers, vol. 1, pp. 120-127, Oct. 1992.
C.A. Hernandez, et al., "Effects of Training Data Distribution on Backpropagation Generalization," 1993 IEEE Int'l. Symp. on Circuits and Systems, vol. 4, pp. 2148-2151, May 1993.
B.-T. Zhang, "An Incremental Learning Algorithm that Optimizes Network Size and Sample Size in One Trial," IEEE World Congress on Computational Intelligence, vol. 1, pp. 215-220, Jun. 1994.
Y. Grandvalet and S. Canu, "Comments on "Noise Injection into Inputs in Back Propagation Learning"," IEEE Trans. on Systems, Man, and Cybernetics, vol. 25(4), pp. 678-681, Apr. 1995.
Khobragade et al, "Connectionist Network for Feature Extraction and Classification of English Alphabetic Characters", IEEE Inter. Conf. on Neural Networks, 1993.
Salicetti et al, "A Neural Predictive Approch for On-Line Cursive Script Recognition", IEEE ICASSP, 1995.
Kim et al, "Multiclass-3D Aircraft Identification and Orientation Estimation Using Multilayer Feed Forward Neural Network", IEEE ICNN 1991.
Nakayama et al, "Handwitten and Digital Character Recognition Using Feature Extracting Neural Network and Modified Selef-Organizing Map", IEEE ICNN 1992.
Spirkovska et al, "An Empirical Comparison of ID3 and HONNs for Distortion Invariant Object Recognition", IEEE Inter. Workshop on Tools for Artificial Intelligence, 1990.
Khotanzad et al, "Classification of Invariant Image Representations Using a Neural Network", IEEE Transaction on Acoustics, Speech, and Signal Processing, vol. 38, No. 6, Jun. 1990.

Affiliated with

Also associated with

Rating

Method for training a statistical classifier with reduced tenden does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Method for training a statistical classifier with reduced tenden, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for training a statistical classifier with reduced tenden will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-255536