Statistical Machine Learning Classifications On Prostate Cancer Dataset

Abstract

Cancer is the second prominent cause of death worldwide. Per annum around 6, 50,000 death cases in this current situation due to Prostate cancer. Need to improve determination the causal factors of prostate cancer. In this research work considers a medical dataset containing clinical information on 100 prostate cancer patients by using the inductive learning algorithms. This research work finds Bayes Net classifier gives an optimal results. The Bayes classifier has highest accuracy level which is 84% of accuracy. The lowest accuracy level is 62% of accuracy which is produced by Naïve Bayes Multinomial Text classifier of Bayes classifier. The Bayes classifier has highest precision level which is 0.85 of precision level. The lowest precision level is 0.62 of precision level which is produced by Naïve Bayes Multinomial Text classifier of Bayes classifier. The Bayes classifier has highest recall level which is 0.84 of recall level. The lowest precision level is 0.62 of recall level which is produced by Naïve Bayes Multinomial Text classifier of Bayes classifier. The Bayes classifier has highest F-Measure level which is 0.84 of F-Measure level. The lowest F-Measure level is 0.76 of F-Measure level which is produced by Naïve Bayes Multinomial Text classifier of Bayes classifier. The Bayes classifier has highest ROC value level which is 0.93 of ROC level. The lowest ROC level is 0.46 of ROC level which is produced by Naïve Bayes Multinomial Text classifier of Bayes classifier. The Bayes classifier has highest PRC value level which is 0.92 of ROC level. The lowest ROC level is 0.51 of PRC level which is produced by Naïve Bayes Multinomial Text classifier of Bayes classifier.