Neural network element with reinforcement/attenuation learning

Details Neural network element with reinforcement/attenuation learning Neural network element with reinforcement/attenuation learning

2005-10-21

2010-02-16

Vincent, David R (Department: 2129)

Data processing: artificial intelligence

Neural network

C706S045000

Reexamination Certificate

active

07664714

ABSTRACT:
A neural network element, outputting an output signal in response to a plurality of input signals, comprises a history memory for accumulating and storing the plurality of input signals in a temporal order as history values. It also includes an output module for outputting the output signal when an internal state exceeds a predetermined threshold value, the internal state being based on a sum of the product of a plurality of input signals and corresponding coupling coefficients. The history values depend on change of the internal state. The neural network element is configured to subtract a predetermined value from the internal state immediately after the output module fires and performs learning for reinforcing or attenuating the coupling coefficient according to the history values after the output module fires.

REFERENCES:
patent: 5113482 (1992-05-01), Lynne
patent: 5222195 (1993-06-01), Alkon et al.
patent: 5704016 (1997-12-01), Shigematsu et al.
patent: 6169981 (2001-01-01), Werbos
patent: 6532454 (2003-03-01), Werbos
patent: 2002/0198854 (2002-12-01), Berenji et al.
patent: 2003/0074338 (2003-04-01), Young et al.
patent: 2003/0149676 (2003-08-01), Kasabov
patent: 2004/0054636 (2004-03-01), Tango-Lowy
patent: 1016981 (2000-07-01), None
patent: 07-182433 (1995-07-01), None
patent: 2000-035956 (2000-02-01), None
Gen Matsumoto and Hiroshi Tsujino, Design of a brain computer using the novel principles of output-driven operation and memory-based architecture: in Cognition and Emotion in the Brain (T. Ono, G. Matsumoto, R. Llinas, A. Berthoz, R. Norgen, H. Nishijo, R. Tamura Eds.), Elsevier, ICS1250, 529-546, Dec. 2003.
Hiroshi Tsujino, Output-driven operation and memory-based architecture principles embedded in a real-world device, Journal of Integrative Neuroscience, 3(2), 133-142, Jun. 2004.
Gen Matsumoto, Hiroshi Tsujino (“Design of a brain computer using the novel principles of output-driven operation and memory-based architecture” International Congress Series 2003.
Barto, A.G. et al., “Neuronlike Adaptive Elements That Can Solve Difficult Learning Control Problems,” IEEE Transactions on Systems, Man and Cybernetics, Sep./Oct. 1983, pp. 834-846, vol. SMC-13, No. 5.
Matsumoto, G. et al., “Design of a Brain Computer Using the Novel Principles of Output-Driven Operation and Memory-Based Architecture” in Cognition and Emotion in the Brain: Selected Topics of the International Symposium on Limbic and Association Cortical Systems (T. Ono, G. Matsumoto, R. Llinas, A. Berthoz, R. Norgren, H. Nishijo, R. Tamura Eds.), held in Toyama, Japan, Oct. 7-12, 2002, International Congress Series No. ICS 1250, 2003, pp. 529-546, Elsevier.
Tsujino, H., “Output-Driven Operation and Memory-Based Architecture Principles Embedded in a Real-World Device,” Journal of Integrative Neuroscience, Jun. 2004, pp. 133-142, vol. 3, No. 2, Imperial College Press.
Tsujino, H., “A System Theory for the Brain-Like Computer,” presented at the Joint 2ndInternational Conference on Soft Computing and Intelligent Systems (SCIS) and 5th International Symposium on Advanced Intelligent Systems (ISIS), Sep. 21-24, 2004, Keio University, Yokohama, Japan.
Tsujino, H., “http://www.brainvision.co.jp/genspage/hrj—tsujino.htm,” 2003, [online] [Retrieved on Sep. 13, 2007] Retrieved from the Internet <URL: http://www.brainvision.co.jp/genspage/hrj—tsujino.htm>.
Machine translation of JP 2000-035956 A (Feb. 2, 2000; Japan Science & Technology Corp.) into English.
Machine translation of JP 07-182433 A (Jul. 21, 1995; Agency of Ind. Science & Technol.) into English.
European Search Report, EP05023033, Jul. 25, 2006.
European Office Action, EP05023033, Feb. 27, 2007.
Communication Pursuant to Article 94(3) EPC (European Patent Office), European Application No. 05023033.3, Jun. 16, 2009, 4 pages.

Affiliated with

Also associated with

Rating

Neural network element with reinforcement/attenuation learning 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 element with reinforcement/attenuation learning, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Neural network element with reinforcement/attenuation learning will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-4156147