PCA-based Feature Transformation for Classification: Issues in Medical Diagnostics

Authors:
Mykola Pechenizkiy;Alexey Tsymbal;Seppo Puuronen
Affiliations:
University of Jyväskylä, Finland;Trinity College Dublin, Ireland;University of Jyväskylä, Finland
Venue:
CBMS '04 Proceedings of the 17th IEEE Symposium on Computer-Based Medical Systems
Year:
2004

Citing 0
Cited 2

Fusing visual and clinical information for lung tissue classification in high-resolution computed tomography

Artificial Intelligence in Medicine
A novel hybrid system with neural networks and hidden markov models in fault diagnosis

MICAI'06 Proceedings of the 5th Mexican international conference on Artificial Intelligence

Quantified Score

Hi-index	0.00

Visualization

Abstract

The goal of this paper is to propose, evaluate, and compare several data mining strategiesthat apply feature transformation for subsequent classification, and to consider theirapplication to medical diagnostics. We (1) briefly consider the necessity of dimensionalityreduction and discuss why feature transformation may work better than feature selection forsome problems; (2) analyze experimentally whether extraction of new components andreplacement of original features by them is better than storing the original features as well; (3)consider how important the use of class information is in the feature extraction process; and(4) discuss some interpretability issues regarding the extracted features.