Prediction In Big Data Context On Scalability Of Machine Learning Models For Breast Cancer

Document Type : Primary Research paper

Authors

1 Assistant Professor, Computer Science and Engineering, Bannari Amman Institute of Technology, Sathyamangalam, Tamilnadu.

2 Associate Professor, Information Technology, K S Rangasamy College of Technology, Tamilnadu, India.

3 Associate Professor, Computer Science and Engineering, Mahendra Institute of Technology, Tamilnadu.

Abstract

Breast cancer is the maximum diagnosed cancer for girls and it is key source of
rising female mortality rates. There is a need to extend the automated prognosis gadget for
early detection of most cancers as the prognosis of ailment physically requires maximum
hours and the lower obtainability of systems. Data mining methods contribute to the
implementation of such a framework. We have used gadget style techniques to obtain
knowledge of the type of benign and malignant tumour in this the device is discovered from
previous statistics and anticipate the current feedback group. This takes is a comparative
look at the application of models that use the Support Vector Machine (SVM) and K
KM(K-MEANS) in the UCI repository dataset. The proficiency of each algorithm is
calculated and compared with the effects of specificity, accuracy, sensitivity, precision, and
Wrong Positive Rate. In Spyder, Scientific Python Development Environment, these
techniques are coded in python and done. Our studies have shown that SVM is very
accurate and excellent for predictive analysis. We assume from our look at SVM as the
right algorithm for prediction and the complete K-MEANS provided well after SVM.

Keywords