JIITA, vol.7 no.2, p.694-708, 2023, DOI: 10.22664/ISITA.2023.7.2.696
Abstract: Heart disease is difficult to detect due to several risk factors, including high blood pressure, cholesterol, and an abnormal pulse rate. Accurate and timely identification of human heart disease can be very helpful in preventing heart failure in its early stages and will improve the patient’s survival. Manual approaches for the identification of heart disease are biased and prone to inter examiner variability. Therefore, detecting heart disease early by utilizing the affluence of high-resolution intensive care records has become a challenging problem. That is why many researchers are trying their best to design a predictive model that can save many lives using data mining. Even though, some Machine Learning (ML) based models are also available, which can reduce the mortality rate, but accuracy is not up to date. According to the recommended study, using a Modified Weighted Empirical Score Optimization (MWESO) with Logistic Regression (LR) algorithm this research identified and predicted human heart disease. Machine learning (ML) algorithms like K-Nearest Neighbourhood (KNN), Support Vector Machine (SVM), Logistic Regression (LR) and Naïve Bayes (NB) have been applied to the heart disease dataset to predict the disease. At first, the LR model was trained. After training, sum of two features decision was combined using a weighted sum optimization. The weights have been assigned to each attribute’s decision probability hence that each attribute’s effect varies in the summation of weighted empirical score that gave the optimized prediction from the final decision score. The datasets were acquired from the heart diseases repositories from Kaggle. The comparative study has proven that the proposed MWESO algorithm with LR is the most suitable model due to its superior prediction capability to other Machine Learning with an accuracy of 90.7% on heart ailments dataset.
Keywords: Heart disease, Prediction, Modified Weighted Empirical Score Optimization, Machine Learning, Classification
Fullpaper: