A Novel Functional Machine Learning Approaches on Prostate Cancer

Abstract

Abstract: Cancer registries are collections of curated data about malignant tumor diseases. The amount of data processed by cancer registries increases every year, making manual registration more and more tedious.This research work finds Bayes Net classifier gives an optimal results. The Sequential Minimal Optimization of functional machine learning approach is having highest accuracy level which is 85% of accuracy level. The Sequential Minimal Optimization of functional machine learning approach is having highest precision level which is 0.85 of precision level. The least precision value is 0.80 of precision value which is having Quadratic Discriminant Analysis of functional machine learning classifier approach. The Sequential Minimal Optimization of functional machine learning approach is having highest recall level which is 0.85 of recall level. The least recall value is 0.79 which is produced by Quadratic Discriminant Analysis functional machine learning classification approach. The Sequential Minimal Optimization of functional machine learning approach is having highest FMeasure level which is 0.85 of F-Measure level. The Fisher’s Discriminant Analysis algorithm of functional machine learning classifier and Linear Discriminant Analysis classification algorithm of functional machine learning classifier are having same receiver operating characteristic curve value which is 0.90 of receiver operating characteristic curve value.The maximum precision recall curve value is 0.90 of precision recall curve value which is produced by Linear Discriminant Analysis of functional machine learning classifier. This system recommends that the Sequential Minimal Optimization of functional machine learning approach produces optimal results compare with other models. Keywords: SMO, functional learning, LDA, QDA, and SDG