A reduced and comprehensible polynomial neural network for classification
β Scribed by B.B. Misra; S. Dehuri; P.K. Dash; G. Panda
- Publisher
- Elsevier Science
- Year
- 2008
- Tongue
- English
- Weight
- 256 KB
- Volume
- 29
- Category
- Article
- ISSN
- 0167-8655
No coin nor oath required. For personal study only.
π SIMILAR VOLUMES
We propose a multigroup classification algorithm based on a hybrid fuzzy neural net framework. A key feature of the approach is the adaptation of membership functions to new data. In this way, learning is reflected in the shape of the membership functions. By defining separate membership functions f
In this paper, a novel modular neural network is proposed to solve multi-class problems with imbalanced training sets. The proposed model can transform an imbalanced classification problem into a set of symmetrical two-class problems, each of which is solved by single neural network with a simple st