Machine Learning-Based Leukemia Classification Using Gene Expression for Accurate Diagnosis
Amena Mahmoud, Kazim Raza Talpur, Shilpa Saini, Bandeh Ali Talpur, Asadullah Shah, John Zaki · 2024
The accurate classification of leukemia is of utmost importance in order to develop effective treatment strategies, given its complex and heterogeneous nature, encompassing numerous subtypes. Given the availability of gene expression data, machine learning algorithms have demonstrated considerable promise in recent years for enhancing the accuracy of leukemia classification. Machine leraning techniques have been used for gene expression dataset in order to classify the existence of Leukemia. A dataset has been taken and thus preprocessed which consist of profiles of gene expression of patients having leukemia. Then the feature selection techniques have been used for classification of informative genes. Various ML techniques have been used here for classification. The proposed techniques is then compared with the existing approaches using the same dataset. It is observed that the proposed model for leukemia classification has an accuracy of 97% using SVM algorithm whereas 94% is using Logistic regression algorithm.