Classification of Clinical Dataset of Cervical Cancer using KNN

Background/Objectives: The primary objective of this paper is to classify the clinical dataset of cervical cancer to identify the stage of cancer which helps in proper treatment of patient suffering from cancer. Methods/Statistical Analysis: This research work basically moves toward the detection of cervical cancer using Pap smear images. Analysis of Pap smear of cervical region is an efficient technique to study any abnormality in cervical cells. The proposed system firstly segment the pap image using Edge Detection to separate the cell nuclei from cytoplasm and background and then extract various features of cervical pap images like area, perimeter, elongation and then these features are normalized using min-max method. After normalization KNN method is used to classify cancer according to its abnormality. Findings: The classification accuracy with 84.3% of maximum performance with no validation and classification accuracy with 82.9% of maximum performance with 5 Fold cross validation is achieved.