NCA classification model using Machine learning for Heart Disease Prediction

  • Ritu Aggarwal, Dr. Suneet Kumar


Now a day a heart disease is the most common disease that is the reason for morality and disability, due to this it is a critical challenge in health care systems. Heart disease is also known as cardiovascular disease. It associates many risk factors and the survival rate is decreasing. This paper proposed a heart disease classification system and comparative analysis between existing systems are implemented. We propose an  classification system  by using approach  Neighborhood Component Analysis (NCA) with the different machine learning classifiers. The experiments results with the NCA obtains the greater performance in is used for selecting the most relevant features.NCA increases classification accuracy and prediction performance. The researchers apply several machine learning classes to analyze large and complex medical data. The aim of this work to propose a classification model related to heart disease by using classifier Random forest, Naïve Bayes, Support vector machine, Decision Tree.  The algorithms to show the best classification and prediction performance by the classifications algorithms. It uses the existing dataset from the Cleveland database that is preprocessed data. The Cleveland dataset was taken from the UCI repository contains 303 instances and 75 attributes. It  has 14 attributes by their features are selected for creating classification system that is selected by training and testing and also performing the comparison between existing systems This paper depicts the higher accuracy score with the multiper classifiers such as Random forest using NCA is 99.342105 %, For  SVM 99.342105% ,for NB 99.342105 % and for DT is 98.026316 %.When Applying NCA and compare its with existing system it provides better and more accurate results.