The Application of Fuzzy Data Processing Technology in the Network Failure Mining

Abstract:

In order to solve the coordinate problem of high redundancy and lack of stability in traditional failure mining knowledge base, this paper presents a fuzzy failure mining algorithm to realize regular acquisition in inconsistent case and purification of learning samples by the comprehensive application of fuzzy data processing technology. With characters of simplified samples, strong adaptability, high failure tolerance and not easy to fall into local minimum point, the algorithm can effectively process the diagnosis and incompatible information in network failure mining. The experiments show that the system implemented by this method improves accuracy and speed of diagnosis and has a certain application value comparing to other similar methods.

• Periodical: Applied Mechanics and Materials (Volumes 380-384)
• Pages: 1073-1076
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.380-384.1073
• Online since: August 2013
• Authors: Lian Tian Li
• Keywords: Computer Network, Failure Diagnosis, Fuzzy Mining

References

[1] J.J. Collins．Fishing for Function in Noise．Nat．1999，402：240-241.

[2] M.I. Dykman，R. Mannella，and P.V.E. McClintock．Giant Nonlinearity in the Low－Frequency Response of a Fluctuating Bistable System．Phys．Rev．E．1993，47：1630-1631.

[3] S. Fauve，F. Heslot. Stochastic Resonance in a Bistable System．Phys．Rev. Ixtt. , 1983, 97A: 6-9.