Filtering non-balanced data using an evolutionary approach

CARBALLIDO, JESSICA A; PONZONI, IGNACIO; CECCHINI, ROCÍO L

LOGIC JOURNAL OF THE IGPL (PRINT)

OXFORD UNIV PRESS

Año: 2022

1367-0751

Matrices that cannot be handled using conventional clustering, regression or classification methods are often found in everybig data research area. In particular, datasets with thousands or millions of rows and less than a hundred columns regularlyappear in biological so-called omic problems. The effectiveness of conventional data analysis approaches is hampered bythis matrix structure, which necessitates some means of reduction. An evolutionary method called PreCLAS is presentedin this article. Its main objective is to find a submatrix with fewer rows that exhibits some group structure. Three stagesof experiments were performed. First, a benchmark dataset was used to assess the correct functionality of the method forclustering purposes. Then, a microarray gene expression data matrix was used to analyze the method’s performance in asimple classification scenario, where differential expression was carried out. Finally, several classification methods werecompared in terms of classification accuracy using an RNA-seq gene expression dataset. Experiments showed that the newevolutionary technique significantly reduces the number of rows in the matrix and intelligently performs unsupervised rowselection, improving classification and clustering methods.