Advanced Machine Learning Frameworks for Cancer Classification Using Genomic Information

Computational advancements have improved patient survival rates and treatment of severe illnesses. Cancer is a hereditary illness caused by mutations and variations in the patient's DNA. Cancer therapy is challenging due to the complexity of the tumor's microenvironment. Patients experiencing the same cancer kind have varying responses to targeted therapy. Clinical trials and conventional drug discovery are complicated and laborious. Thus, researchers strive to develop effective treatments for severe disorders. The abundance of internet data sources in oncology and pharmacogenomics has encouraged research in this discipline. Recent advances in data mining including machine learning are enhancing analysis based on data. The following section introduces a cancer classification method that uses multi-objective meta-heuristics and machine learning techniques to predict novel and relevant markers. HMOSHSSA, a hybrid choice of genes method, uses MOSHO and Salp Swarm method. The proposed composite algorithm combines SSA and MOSHO characteristics for mining and extraction. The suggested strategy is compared to a few state-of-the-art methods. The precision of classification and gene count are used to assess the proposed approach. Comparing findings shows that the recommended strategy outperforms competitors. The SVM and KNN classification method for leukemia dataset identifies four genes with 100% accuracy. Existing leukemia dataset prediction methods are less accurate.