Published Paper Details:

In cancer diagnosis and drug creation, the recognition of various tumor types is important. The majority of previous cancer classification studies, on the other hand, has been clinically based and has limited diagnostic potential. The keys to solving the fundamental problems related to cancer detection and drug development are known to be found in cancer classification using gene expression data. Researchers have begun to investigate the possibilities of cancer classification using gene expression data as a result of the abundance of gene expression data. Because of the dataset's scale, visual analysis cannot be used to determine the number of missing values. This paper depicts the novel approach of pre-processing phases of gene expression data using gene pre-processor, with a focus on the filtering and normalization steps that can effectively deal with missing values in a dataset with thousands of rows and columns for cancer type classification. The experimental analysis shows that the proposed approach of gene pre-processor produces efficient results having reduced the size of the dataset with less elapsed time.