ABSTRACT
Numerous examples abound on the use of specific algorithms or techniques to analyze the results of specific array experimental data and several provide comparative studies.
Given such a dataset, which algorithm should we select in mining gene expression data? Are we looking to cluster? To classify? To identify associations or other relationships? Are we looking to build a predictor? To analyze time series? Are we verifying a hypothesis? All these questions are relevant and each is repeated numerous times in every area of bioinformatics and cheminformatics. The large number of available techniques to choose from does not simplify the problem.
