id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_lgr5suct7rhcdlkzwade3g4fhm	Tomasz Latkowski	Data mining for feature selection in gene expression autism data	2015	6	.pdf	application/pdf	5483	547	66	Feature selection methods in application to gene expression: The paper presents the application of several different feature selection methods for recognizing the most significant genes and gene sequences (treated as features) stored in dataset of gene expression microarray related to autism. In the next step fusion of the most relevant features selected by different The optimal number of features has been defined as the set providing the best clustering purity. Keywords: feature selection, gene expression microarrays, clustering, autism. Present data mining methods allow to look at the gene application of feature selection methods allows to identify a different feature selection methods will be examined and values of the randomly selected gene for the autism and randomly selected gene representing two classes (autism and The other statistical feature selection method applied in mean value of the feature for the kth class data, σ2(f) is a set of genes selected individually by different methods and after 	./cache/work_lgr5suct7rhcdlkzwade3g4fhm.pdf	./txt/work_lgr5suct7rhcdlkzwade3g4fhm.txt