A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Unbounded Fuzzy Hypersphere Neural Network Classifier
https://orcid.org/http://orcid.org/0000-0003-2641-4589
This paper presents the supervised classifier called unbounded fuzzy hypersphere neural network (UFHSNN) model. The basic fuzzy min-max neural network (FMMN), fuzzy hypersphere neural network (FHSNN) and many more its variants use expansion parameter to tune the model. Such tuned model gives good classification accuracy with minimum number of hyperboxes, but always needs to take multiple iterations to find optimal value of expansion parameter and train the model. However, in the proposed model, use of expansion parameter is removed, due to which the model can be trained in a single iteration. The different datasets are applied for verification of the model; also, outcomes are compared with some current FMMN variations. The analysis of outcomes shows that the presented model gives magnificent accuracy with reduced computational complexity.
Unbounded Fuzzy Hypersphere Neural Network Classifier
https://orcid.org/http://orcid.org/0000-0003-2641-4589
This paper presents the supervised classifier called unbounded fuzzy hypersphere neural network (UFHSNN) model. The basic fuzzy min-max neural network (FMMN), fuzzy hypersphere neural network (FHSNN) and many more its variants use expansion parameter to tune the model. Such tuned model gives good classification accuracy with minimum number of hyperboxes, but always needs to take multiple iterations to find optimal value of expansion parameter and train the model. However, in the proposed model, use of expansion parameter is removed, due to which the model can be trained in a single iteration. The different datasets are applied for verification of the model; also, outcomes are compared with some current FMMN variations. The analysis of outcomes shows that the presented model gives magnificent accuracy with reduced computational complexity.
Unbounded Fuzzy Hypersphere Neural Network Classifier
https://orcid.org/http://orcid.org/0000-0003-2641-4589
This paper presents the supervised classifier called unbounded fuzzy hypersphere neural network (UFHSNN) model. The basic fuzzy min-max neural network (FMMN), fuzzy hypersphere neural network (FHSNN) and many more its variants use expansion parameter to tune the model. Such tuned model gives good classification accuracy with minimum number of hyperboxes, but always needs to take multiple iterations to find optimal value of expansion parameter and train the model. However, in the proposed model, use of expansion parameter is removed, due to which the model can be trained in a single iteration. The different datasets are applied for verification of the model; also, outcomes are compared with some current FMMN variations. The analysis of outcomes shows that the presented model gives magnificent accuracy with reduced computational complexity.
Unbounded Fuzzy Hypersphere Neural Network Classifier
J. Inst. Eng. India Ser. B
Mahindrakar, M. S. (author) / Kulkarni, U. V. (author)
Journal of The Institution of Engineers (India): Series B ; 103 ; 1335-1343
2022-08-01
9 pages
Article (Journal)
Electronic Resource
English
Fuzzy ARTMAP - A Neural Classifier for Multispectral Image Classification
| British Library Conference Proceedings | 1997
The Possibility of Architectural Designing through the Medium of 4 Dimensional Hypersphere
| British Library Online Contents | 1995
| British Library Conference Proceedings | 2000
Fuzzy Logic Distress Classifier for Pavement Imaging
| British Library Conference Proceedings | 1997
Fuzzy Logic Distress Classifier for Pavement Imaging
| British Library Conference Proceedings | 1998