Heart disease classification using hybrid ML schemes and optimization tactics in healthcare.

Problem: Heart disease remains a major contributor to mortality worldwide, and early diagnosis is crucial for treatment. Traditional diagnostic tactics often face challenges regarding accuracy and efficacy. With the rise of ML and decision-making approaches, there is a rising interest in developing automated systems to aid in heart disease detection.

Aim: This investigation tries to boost the accuracy of heart disease classification by integrating advanced optimization schemes with machine learning (ML) schemes, specifically the Random Forest Classifier (RFC) and Gaussian Process Classifier, to boost diagnostic performance for heart disease projection.

Tactics: Four hybrid schemes were developed by integrating the Golf optimization algorithm (GOA) and Alibaba optimization algorithm with the RF and Gaussian process classifiers. The hybrid schemes were trained and evaluated on a comprehensive database of clinical factors related to heart disease. Data preprocessing included random permutation, missing value imputation, and a 70-30 split into training and test sets.

Outcomes: In the recommended schemes, the RF with GOA had the maximum classification accuracy of 95.38%, which is 4.33% higher than the individual RF model. This is greater than the comparative study's best accuracy, which is approximately 92.32%, and demonstrates the efficacy of RFGO in classifying heart disease patients with high accuracy.

Conclusion: The RF with GOA significantly improves the accuracy of heart disease classification, illustrating its strong application as a high-performance tool for use in decision support systems in managing cardiovascular health. The results indicate the significance of implementing optimization tactics in ML schemes to boost healthcare diagnostic capabilities.