Experimental investigation of three machine learning algorithms for ITS dataset

The present article is devoted to experimental investigation of the performance of three machine learning algorithms for ITS dataset in their abil- ity to achieve agreement with classes published in the biological literature be- fore. The ITS dataset consists of nuclear ribosomal DNA sequences, where rather sophisticated alignment scores have to be used as a measure of distance. These scores do not form a Minkowski metric and the sequences cannot be re- garded as points in a finite dimensional space. This is why it is necessary to de- velop novel machine learning approaches to the analysis of datasets of this sort. This paper introduces a k-committees classifier and compares it with the dis- crete k-means and Nearest Neighbour classifiers. It turns out that all three machine learning algorithms are e␣cient and can be used to automate future biologically significant classifications for datasets of this kind. A simplified ver- sion of a synthetic dataset, where the k-committees classifier outperforms k-means and Nearest Neighbour classifiers, is also presented.