Prediction and categorization of ovarian cancer from big data using data mining techniques

Abstract

Data mining is the retrieval of potentially useful information from massive amount of data. The role of information and communication technology is crucial in health care industry. Errors in disease diagnostics will hike the mortality rate of patients and also create unnecessary complications for the stake holders. There is an increased need in creating proper communication channel among patients, family members and clinicians at right time. Data mining techniques are deployed in oncology, pathology and diagnosis of rare diseases for pattern matching. The machine learning systems help to save the time which is more precious in the health care sector. This research work focus on creating a framework for addressing the ovarian cancer in oncology.Ovarian cancer is the most frequent and common found disease on women. This cancer would spread quickly throughout the entire body that might cause the human loss. Ovarian cancer needs to diagnose early to prevent the human loss and treat the women. In this research work, early detection of ovarian cancer presence on women are concentrated. This is done by introducing the several research methods which could lead to accurate and early diagnosis of ovarian cancer presence. Three novel methods are deployed for the problem subsequently and results are presented. In the first research method, Hybridized Bacterial Foraging with Particle Swarm and Multi Kernel Support Vector Machine Approach (OCD_HBFMPSO_MKSVM) is introduced for ovarian cancer disease prediction which predict cancer occurrence by examining the gene expression information gathered from the women newline