IEEE Transactions on Neural Networks
The geometrical learning of binary neural networks
IEEE Transactions on Neural Networks
Semi-supervised projected model-based clustering
Data Mining and Knowledge Discovery
| Hi-index | 0.00 |
In this paper, we propose a novel fast training algorithm called Constructive Semi-Supervised Classification Algorithm (CS-SCA) for neural network construction based on the concept of geometrical expansion. Parameters are updated according to the geometrical location of the training samples in the input space, and each sample in the training set is learned only once. It's a semi-supervised based approach, the training samples are semi-labeled i.e. for some samples, labels are known and for some samples, data labels are not known. The method starts with clustering, which is done by using the concept of geometrical expansion. In clustering process various clusters are formed. The clusters are visualizes in terms of hyperspheres. Once clustering process over labeling of hyperspheres is done, in which class is assigned to each hypersphere for classifying the multi-dimensional data. This constructive learning avoids blind selection of neural network structure. The method proposes here is exhaustively tested with different benchmark datasets and it is found that, on increasing value of training parameters number of hidden neurons and training time both are getting decrease. Through our experimental work we conclude that CS-SCA result in simple neural network structure by less training time.