Classification of microarrays; synergistic effects between normalization, gene selection and machine learning.

Machine learning and gene selection (Figure 9 ) A clear trend is that the T-test is the best gene selection partner to all machine learning methods, although the actual improvement sometimes is rather small.

also in 19,460 other papers

The permutation tests show that shuffled data sets result in adjusted error rates centered close to 1.0 and that our reported error rates for the original data sets are highly significant with the exception of Sørlie.

also in 132,142 other papers