Dynamic Growing Self-organizing Neural Network for Clustering

Authors:
Daxin Tian;Yueou Ren;Qiuju Li
Affiliations:
School of Computer Science and Technology, Tianjin University, China 300072;Department of Electronic Engineering, Armor Technique Institute of PLA, China 130117;Department of Electronic Engineering, Armor Technique Institute of PLA, China 130117
Venue:
ADMA '08 Proceedings of the 4th international conference on Advanced Data Mining and Applications
Year:
2008

Citing 5
Cited 1

A Game-Theoretic Adaptive Categorization Mechanism for ART-Type Networks

ICANN '01 Proceedings of the International Conference on Artificial Neural Networks
Analyses on the Generalised Lotto-Type Competitive Learning

IDEAL '00 Proceedings of the Second International Conference on Intelligent Data Engineering and Automated Learning, Data Mining, Financial Engineering, and Intelligent Agents
A self-organising network that grows when required

Neural Networks - New developments in self-organizing maps
Adaptive double self-organizing maps for clustering gene expression profiles

Neural Networks - 2003 Special issue: Advances in neural networks research — IJCNN'03
A feature extraction unsupervised neural network for an environmental data set

Neural Networks - 2003 Special issue: Neural network analysis of complex scientific data: Astronomy and geosciences

Self adaptive growing neural network classifier for faults detection and diagnosis

Neurocomputing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Neural Networks have been widely used in the field of intelligent information processing such as classification, clustering, prediction, and recognition. Unsupervised learning is the main method to collect and find features from large unlabeled data. In this paper a new unsupervised learning clustering neuron network--Dynamic Growing Self-organizing Neuron Network (DGSNN) is presented. It uses a new competitive learning rule--Improved Winner-Take-All (IWTA) and adds new neurons when it is necessary. The advantage of DGSNN is that it overcomes the usual problems of other clustering methods: dead units and prior knowledge of the number of clusters. In the experiments, DGSNN is applied to clustering tasks to check its ability and is compared with other clustering algorithms RPCL and WTA. The results show that DGSNN performs accurately and efficiently.