Method for training neural networks

Details Method for training neural networks Method for training neural networks

2008-07-29

2008-07-29

Starks, Jr., Wilbert L (Department: 2129)

Data processing: artificial intelligence

Neural network

Learning task

C706S012000, C706S015000, C706S020000, C706S045000, C600S409000

Reexamination Certificate

active

07406451

ABSTRACT:
The invention relates to a system and a method for training a number of neural networks, by determining a first training data record, wherein the training data have a particular accuracy, generating a number of second data training records by perturbing the first training data record with a random variable, and training each of the neural networks with one of the training data records. A prognosis and an estimation of the prognosis error can be carried out by means of such a system.

REFERENCES:
patent: 5417211 (1995-05-01), Abraham-Fuchs et al.
patent: 5699487 (1997-12-01), Richardson
patent: 5704012 (1997-12-01), Bigus
patent: 5745652 (1998-04-01), Bigus
patent: 5826249 (1998-10-01), Skeirik
patent: 6282529 (2001-08-01), Neuneier et al.
patent: 6523016 (2003-02-01), Michalski
patent: WO 93/00652 (1993-01-01), None
patent: 9810352 (1998-03-01), None
patent: 9834177 (1998-08-01), None
patent: WO 98/3477 (1998-08-01), None
Yan B et al “Application of probabilistic neural network and static test data to the classification of the bridge damage patterns” Proceedings of the SPIE- the International Society for Optical Engineering SPIE-Int. Soc. Opt. Eng USA Mar. 3-6, 2003 BD. 505, Marz 2003, seiten 606-617 XP 002347955 ISSN: 0277-786X.
Holmstrom L et al: “Using additive noise in back-propagation training” IEEE Transactions on Neural Networks, USA, Bd. 3 Nr. 1, Jan. 1992 Seiten 24-38, XP 00262184 ISSN:1045-9227.
Banfu Yan et al., Application of Probabilistic Neural Network and Static Test DAta to the Classification of BRidge Damage Patterns, Smart Structures and Materials 2003: Smart Systems and Nondestructive Evaluation for Civil Infrastructures, 2003, pp. 606-617, vol. 5057.

Affiliated with

Also associated with

Rating

Method for training neural networks 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 neural networks, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for training neural networks will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2808587