Computational Learning Model for Prediction of Heart Disease Using Machine Learning Based on a New Regularizer.

Area of Science:

• Cardiology
• Computational Biology
• Machine Learning

Background:

• Heart diseases are complex and heterogeneous, necessitating accurate early diagnosis and prognosis.
• Effective clinical management of heart disease patients relies on timely and precise predictive tools.

Purpose of the Study:

• To introduce a novel computational model for the early prediction of heart disease.
• To evaluate the efficacy of the proposed model against existing methods.

Main Methods:

• Development of a new regularization technique for predictive modeling.
• Implementation of a computational model named RSD-ANN, incorporating weight decay based on standard deviation and parent comparison.
• Utilizing tenfold cross-validation and holdout methods for performance evaluation.

Main Results:

• The proposed RSD-ANN model demonstrated superior performance compared to existing methods.
• An average validation accuracy of 96.30% was achieved using both tenfold cross-validation and holdout validation strategies.

Conclusions:

• The RSD-ANN model offers a significant advancement in the early prediction of heart disease.
• This computational approach holds promise for improving the clinical management and prognosis of patients with heart conditions.