Comparison of hierarchic agglomerative clustering methods for document retrieval
The Computer Journal
Principal component neural networks: theory and applications
Principal component neural networks: theory and applications
New approximations of differential entropy for independent component analysis and projection pursuit
NIPS '97 Proceedings of the 1997 conference on Advances in neural information processing systems 10
A Fast and High Quality Multilevel Scheme for Partitioning Irregular Graphs
SIAM Journal on Scientific Computing
Natural gradient learning for over- and under-complete bases in ICA
Neural Computation
Concept decompositions for large sparse text data using clustering
Machine Learning
Flexible Independent Component Analysis
Journal of VLSI Signal Processing Systems
Fast Independent Component Analysis in Kernel Feature Spaces
SOFSEM '01 Proceedings of the 28th Conference on Current Trends in Theory and Practice of Informatics Piestany: Theory and Practice of Informatics
A methodology for analyzing SAGE libraries for cancer profiling
ACM Transactions on Information Systems (TOIS)
Equivariant adaptive source separation
IEEE Transactions on Signal Processing
Frequent variable sets based clustering for artificial neural networks particle classification
APWeb/WAIM'07 Proceedings of the joint 9th Asia-Pacific web and 8th international conference on web-age information management conference on Advances in data and web management
| Hi-index | 0.00 |
We present the use of KICA to perform clustering of gene expression data. Comparison experiments between KICA and two other methods, PCA and ICA, are performed. Three clustering algorithms, including weighted graph partitioning, k-means and agglomerative hierarchical clustering, and two similarity measures, including Euclidean and Pearson correlation, are also evaluated. The results indicate that KICA is an efficient feature extraction approach for gene expression data clustering. Our empirical study showed that clustering with the components instead of the original variables does improve cluster quality. In particular, the first few components by KICA capture most of the cluster structure. We also showed that clustering with components has different impact on different algorithms and different similarity metrics. Overall, we would recommend KICA before clustering gene expression data.