ML Classifier Comparative Performance Analysis of Prediction on Cervical Cancer
Md. Ashfaqul Haque, Israt Jahan Dristy, Shihab Sharar, Annajiat Alim Rasel · 2021 International Conference on Electronics, Communications and Information Technology (ICECIT) · 2021
Identification of cancer at an early stage in treating patients has been worthwhile. Cervical cancer occurs by alteration of cells connecting the uterus and vulva to anomalous changed state which can culminate in infecting deeper tissue. By making an early prediction on the result of the biopsy test, it may come to the verge of a cure. For making prognostication on biopsy test we used a dataset provided by the hospital Universitario de Caracas. Analytical and statistical survey on the dataset has led us to a conjecture of making predictions on biopsy using the following six classifiers which are LR, random forest, NB, SVM, SGD, and kNN, and finally done by comparing performance analysis. Among them, multinomial NB has stood out as the best performer with an accuracy of 97.76 percent.