The Density-Based Agglomerative Information Bottleneck

Authors:
Yongli Ren;Yangdong Ye;Gang Li
Affiliations:
School of Information Engineering, Zhengzhou University, Zhengzhou, China;School of Information Engineering, Zhengzhou University, Zhengzhou, China;School of Engineering and Information Technology, Deakin University, Australia Vic 3125
Venue:
PRICAI '08 Proceedings of the 10th Pacific Rim International Conference on Artificial Intelligence: Trends in Artificial Intelligence
Year:
2008

Citing 3
Cited 0

Document clustering using word clusters via the information bottleneck method

SIGIR '00 Proceedings of the 23rd annual international ACM SIGIR conference on Research and development in information retrieval
Unsupervised document classification using sequential information maximization

SIGIR '02 Proceedings of the 25th annual international ACM SIGIR conference on Research and development in information retrieval
Unsupervised image-set clustering using an information theoretic framework

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

The Information Bottleneck method aims to extract a compact representation which preserves the maximum relevant information. The sub-optimality in agglomerative Information Bottleneck (aIB) algorithm restricts the applications of Information Bottleneck method. In this paper, the concept of density-based chains is adopted to evaluate the information loss among the neighbors of an element, rather than the information loss between pairs of elements. The DaIB algorithm is then presented to alleviate the sub-optimality problem in aIB while simultaneously keeping the useful hierarchical clustering tree-structure. The experiment results on the benchmark data sets show that the DaIB algorithm can get more relevant information and higher precision than aIB algorithm, and the paired t-test indicates that these improvements are statistically significant.