An Optimized Classification Approach with Fused Feature Selection Process for Diagnosing Gynecological Cancers

Abstract

Gynecological cancer consists of those cancers that originate from the female newlinereproductive system, primarily with ovary, cervix, endometrium, vulva or vagina. In newlinedeveloping countries like India, ovarian and cervical cancers are identified to be the majorly newlineoccurring gynecological cancer types that affect women of various age groups and most of newlinethese complications are related to the human papillomavirus (HPV) infection. Many risk newlinefactors are present which are associated with various types of gynecological cancers. As newlinecervical and ovarian cancers have been recognized as critical types of gynecological newlinecancers, identifying the effect of various test variables of these cancer types is significant newlinefor diagnosing the patients and classifying the cancer stages centered on the results. This newlineproposed research aimed at achieving a deeper insight by the application of Machine newlineLearning practices to examine risk features that are significant for diagnosis and staging newlineclassifications of gynecological cancers that are more critical. newlineFirstly, this proposed research aims in reviewing on various tools, techniques, and newlineframeworks of Machine Learning for optimal diagnosis and classifications of gynecological newlinecancers with current studies. Secondly, the proposed study aimed to put forward a newlinecomprehensive structure or framework using feature selection processes that are fused to newlineachieve enhanced subgroup of features for efficient classification method. Next this newlineresearch intends to implement and evaluate the planned framework for cervical cancer newlinediagnosis and classification. Subsequently it aims for an enhanced categorization approach newlineutilizing the technique of fused feature/attribute selection by the combination of filter in newlineaddition to wrapper-based feature selection approaches for staging classifications of newlinegynecological cancer patient s data. Then this proposed research aims to associate the newlineperformance of proposed model in finding the optimal feature subset for cervical cancer newlinediagnosis and staging classificat