Abstract | ||
---|---|---|
This paper describes Modular Fuzzy Hypersphere Neural Network (MFHSNN) with its learning algorithm, which is an extension of Fuzzy Hypersphere Neural Network (FHSNN) proposed by Kulkarni and Sontakke [2001]. The MFHSNN offers higher degree of parallelism. Each module in MFHSNN is exposed to the patterns of only one class and trained without overlap test and removal, unlike in FHSNN, leading to reduction in training time. Hence, each module captures peculiarity of only one particular class and due to decrease in training time the algorithm can be used for voluminous realistic database, where new patterns can be added on fly. The MFHSNN is found superior than FHSNN in terms of generalization and training time with equivalent testing time. |
Year | DOI | Venue |
---|---|---|
2003 | 10.1109/FUZZ.2003.1209367 | Fuzzy Systems, 2003. FUZZ '03. The 12th IEEE International Conference |
Keywords | Field | DocType |
fuzzy neural nets,generalisation (artificial intelligence),handwritten character recognition,learning (artificial intelligence),query processing,Fisher Iris database,equivalent testing time,fuzzy membership function,generalization,higher degree of parallelism,learning algorithm,machine-learning databases,modular fuzzy hypersphere neural network,rotation invariant handwritten character recognition,training time reduction,voluminous realistic database | Pattern recognition,Computer science,Degree of parallelism,Fuzzy logic,Hypersphere,Network topology,Fuzzy set,Unsupervised learning,Artificial intelligence,Modular design,Artificial neural network,Machine learning | Conference |
Volume | ISSN | ISBN |
1 | 1098-7584 | 0-7803-7810-5 |
Citations | PageRank | References |
2 | 0.39 | 7 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Pradeep M. Patil | 1 | 86 | 6.66 |
Kulkarni, U.V. | 2 | 8 | 1.32 |
Sontakke, T.R. | 3 | 5 | 0.81 |