An extensive experimental analysis for heart disease prediction using artificial intelligence techniques.

Area of Science:

• Cardiology
• Artificial Intelligence
• Machine Learning

Background:

• Heart disease is a leading global cause of mortality.
• Early and accurate detection of heart disease is critical for effective treatment and prevention.
• Artificial intelligence (AI) offers promising solutions for healthcare, particularly in disease prediction.

Purpose of the Study:

• To investigate and compare the effectiveness of various machine learning models for heart disease prediction.
• To identify the optimal model for accurate and reliable heart disease detection.

Main Methods:

• The study evaluated 11 feature selection techniques and 21 distinct classification algorithms.
• Feature selection methods included Information Gain, Chi-Square Test, FDA, PCA, and others.
• Classifiers encompassed Logistic Regression, SVM, Random Forest, XGBoost, various neural networks (CNN, RNN, LSTM), and hybrid models.

Main Results:

• XGBoost significantly outperformed all other evaluated models.
• XGBoost achieved an accuracy of 0.97, precision of 0.97, sensitivity of 0.98, specificity of 0.98, F1 score of 0.98, and AUC of 0.98.
• These results indicate XGBoost's high efficacy in heart disease prediction.

Conclusions:

• The research highlights the potential of AI, specifically XGBoost, in improving heart disease prediction accuracy.
• The findings suggest that XGBoost can be a valuable tool for early detection and personalized healthcare recommendations.
• Further experimentation with diverse models is crucial for advancing AI in cardiovascular health.