Cancer Class Prediction: An Empirical Analysis of Gene Expression Data

M. Shin (Korea), A.L. Goel, and H. Lim (USA)


Microarray data analysis; Gene expression; Radial basis function; Cancer classification; Cancer class prediction.


We describe a method for binary cancer classification based on gene expression data from DNA microarray hybridization experiments. The underlying architecture of this method is the gaussian basis function model but our model development process is substantially different from and superior to the current approaches. Cancer class prediction is crucial to its treatment and developing an analytical approach for classification based upon the microarray expression is an important problem. A generic approach to classifying two types of acute leukemia based on gene expression monitoring by DNA micro arrays was originally pioneered by Golub [1]. In this paper, we employ their data sets to illustrate the use of our method, provide insights into the classifier development process and undertake sensitivity analyses for selecting an appropriate cancer class prediction model.

Important Links:

Go Back