Neural Network Reliability

Chapter

Abstract

The first attempts to estimate the neural network functional reliability were experimental [15-1] or qualitative [15-2]. The qualitative estimations showed that the neuron-like elements are characterized by the logical redundancy [15-2], [15-3], i.e., the failures of some elements do not result in the errors at the neural network output.

Keywords

Neural Network Weighting Coefficient Parametrical Reliability Multilayer Neural Network Optimal Realization 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
This is a preview of subscription content, log in to check access.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Literature

  1. [15-1]
    Coates CL, Lewis PM (1964) DONUT-a threshold gate computer. IEEE Trans. ELEMENT. Comp.; No. 3, vol. EC-13Google Scholar
  2. [15-2]
    Abbakumov IS, Chernyshv NA (1979) Majority redundancy systems with readjustable structure. Automatics and computer facilities 3:31–36Google Scholar
  3. [15-3]
    Blyum M, Onesto M, Verbik L (1966) Acceptable neuron errors for neural network failure-free performance. In: Methods of redundancy introduction for computer systems. Pougatchev VS (ed) — Moscow: Sov. Radio, p 84–87Google Scholar
  4. [15-4]
    Potapov VI, Palyanov IA (1972) To the functional reliability estimation of the redundancy readjustable homogeneous computer structure. In: Computer facilities in the in the systems of flight vehicle control, part II, vyp. 23, MoscowGoogle Scholar
  5. [15-5]
    Potapov VI (1977) Analysis and synthesis of highly reliable digital and computer threshold unit logical structures. Novosibirsk, p 80Google Scholar
  6. [15-6]
    Potapov VI (1968) Functional reliability of the formal neuron networks. Automatics and computer facilities 1:37–43Google Scholar
  7. [15-7]
    Maitra KK (1966) Reliable automats synthesis and neuron circuit stability. Bionics, parts II and III. Kiev: KVIITRU, p 37–43Google Scholar
  8. [15-8]
    Lopin VN (1975) About reliability of the controlled threshold element network under constraints upon its complexity. In: Adaptive control systems. Kiev, p 91–97Google Scholar
  9. [15-9]
    Serapinas KL, Jukauskas KP (1969) Threshold element reliability (1. Probabilistic estimation of the relay part of the threshold element). Trudy AN LitCCP, ser. B, 2(57):159–162Google Scholar
  10. [15-10]
    Jukauskas KP, Serapinas KL (1969) Threshold element reliability (2. External noise influence.) Trudy AN LitCCP, ser. B, 4(59):213–216Google Scholar
  11. [15-11]
    Jukauskas KP, Serapinas KL (1970) Threshold element reliability (3. Generalized VRF for the group of threshold elements). Trudy AN LitCCP, ser. B, 3(62):153–157Google Scholar
  12. [15-12]
    Jukauskas KP, Serapinas KL (1971) Threshold element reliability (5. Determination of the average threshold element error taking into account the input signals and weighting coefficient parameter spread). Trudy AN LitCCP, ser. B, 1(64):231–236Google Scholar
  13. [15-13]
    Gill A. (1974) Linear sequential machines. Moscow: Mir, p. 287Google Scholar
  14. [15-14]
    Galushkin AI (1974) Synthesis of multilayer pattern recognition systems. Moscow: Energiya, 367 pGoogle Scholar
  15. [15-15]
    Palianov IA, Potapov VI (1977) Failure diagnostics and synthesis of digital structures based on the threshold logical units. Novosibirsk, p 78Google Scholar
  16. [15-16]
    Neumann J (1955) Probabilistic logic and synthesis of reliable organisms on the basis of unreliable components. In: Automats, Moscow: Foreign litr., p 68–138Google Scholar
  17. [15-17]
    Pirs W (1968) Design of reliable computers. Moscow: Mir, p 270Google Scholar
  18. [15-18]
    Mour E, Shennon K (1960) Reliable schemes of unreliable relays. In: Cybernetics collection. Moscow: Foreign litr., vyp. 1, p 109–149Google Scholar
  19. [15-19]
    Trayon J (1966) Quadruplicate logic. In: Methods of redundancy introduction for computer systems. Moscow: Sov. RadioGoogle Scholar
  20. [15-20]
    Elias P (1958) IBM Journal of research and development, No. 3, 1958, posterior probability, pp 346–353MathSciNetCrossRefGoogle Scholar
  21. [15-21]
    Winograd S, Cowan JD (1963) Reliable computation in the presence of noise, M.I.T. Press, Cambridge, Mass.MATHGoogle Scholar
  22. [15-22]
    Zrelova TI (1978) Review of methods for computer system reliability improvement. In: Computers and readjustable structure systems. Ser. Cybernetics problems, p 152–163Google Scholar
  23. [15-23]
    Malev VA (1978) Structure redundancy in the logical devices. Moscow: Sviaz, p 192Google Scholar
  24. [15-24]
    Bennets RG (1978) Designing reliable computer systems. The fault-tolerant approach-I. Electron and Power, posterior probability, pp 846–851Google Scholar
  25. [15-25]
    Chernyshov Yu A, Abbakoumov IS (1979) Computing and design of computer devices with passive reservation. Moscow: Energiya, p 119Google Scholar
  26. [15-26]
    Mathur FP, de Sousa PT (1975) Reliability modeling and analysis of general modular redundant systems. IEEE Trans. Reliability, vol. R-24, posterior probability, p 296–299CrossRefGoogle Scholar
  27. [15-27]
    Mathur FP, Avizienis A (1970) Reliability analysis and architecture of a highly redundant digital system. Generalized triple modular redundancy with self repair. Proc. SJCC, vol. 26, posterior probability, p 375–383Google Scholar
  28. [15-28]
    Pakoulev NI, Ukhanov VM, Chernyshov PN (1974) Majority principle of design of units and devices of digital computers. Moscow: Sov. Radio, p. 183Google Scholar
  29. [15-29]
    Losev VV (1971) Restoring organs on the basis of majority elements. Izv. AN SSSR, Technology cybernetics 2:116–122MathSciNetGoogle Scholar
  30. [15-30]
    Fomin Yu I (1980) About restoring organs realizing majority voting. Electronic modeling 2:53–60Google Scholar
  31. [15-31]
    Fomin Yu I (1980) Program for localization of parametrical failures. Annotated checklist of new receipts. MosFAP ASU 2:10Google Scholar
  32. [15-32]
    Fomin Yu I (1980) Program for generation of minimum fault detection test. Ibidem, p 11Google Scholar
  33. [15-33]
    Fomin Yu I (1980) Program for investigation of catastrophic reliability. Ibidem, p 11–12Google Scholar
  34. [15-34]
    Fomin Yu I (1980) Program for investigation of parametrical reliability. Ibidem, p 12Google Scholar
  35. [15-35]
    Fomin Yu I (1980) Program for the calculation of the failure probability of logical device consisting of duplex identical blocks with the network of majority elements at the input. Ibidem, p 13–14Google Scholar
  36. [15-36]
    Fomin Yu I (1980) Program for the calculation of the failure probability of logical device consisting of duplex identical blocks with the network of majority elements at the output. Ibidem, p 13Google Scholar
  37. [15-37]
    Potapov VI, Palianov IA (1973) Design of fault detection tests for the threshold elements. Izv AN SSSR, Technology cybernetics 4:140–146Google Scholar
  38. [15-38]
    Fomin Yu I, Galushkin AI (1982) Majority voting and restoring organs for its implementation. Cybernetics and computer facilities. Vyp. 55, Kiev, Naukova Dumka, p 91–97MATHGoogle Scholar
  39. [15-39]
    Mkrtchan SO (1977) Design of computer logical devices on the basis of neuron elements. Moscow: Energiya, p 199Google Scholar
  40. [15-40]
    Galushkin AI, Fomin Yu I (1979) About optimality of restoring organs realizing majority voting. Tekhnika sredstv sviazi, ser. ASU 3:56–61Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2007

Personalised recommendations