A Data Mining Model To Predict Breast Cancer Using Improved Feature Selection Method On Real Time Data
Amrita Sanjay, Hari Nair, Sruthy Murali, K S Krishnaveni · 2018
Medical data mining models are predominant in medical data analysis. In our paper, we have applied the Naive Bayes (NB) classifier algorithm, along with a new improved feature selection method, towards an accurate detection of breast cancer, in its early stages. Performance of the NB classifier was seen to be noticably enhanced when irrelevant features were screened out from the modeling process. Empirical verification reaffirm that our hybrid feature-selection approach, using minimal set of attributes, outperforms the results obtained from solitary feature selection methods.