A hybrid machine learning approach using particle swarm optimization for cardiac arrhythmia classification.

Area of Science:

• Cardiology
• Computer Science
• Artificial Intelligence

Background:

• Accurate cardiac arrhythmia identification is crucial for patient care.
• Machine learning (ML) shows promise for arrhythmia classification, but requires hyperparameter tuning.
• Optimizing ML models is key to improving diagnostic accuracy.

Purpose of the Study:

• To develop and evaluate a novel hybrid strategy for cardiac arrhythmia classification.
• To enhance the predictive performance of various ML algorithms using Particle Swarm Optimization (PSO).
• To assess the efficacy of PSO-optimized ML models on the UCI cardiac arrhythmia dataset.

Main Methods:

• A synergistic approach combining PSO with ML algorithms (Logistic Regression, Linear Discriminant Analysis, Gaussian Naive Bayes, Decision Tree, XGBoost Classifier).
• Implementation and validation of models on the UCI cardiac arrhythmia dataset using Stratify K-Fold.
• Hyperparameter optimization of ML models through PSO.

Main Results:

• Hybrid models significantly outperformed unoptimized counterparts.
• PSO-optimized XGBoost Classifier (Model 5) achieved 95.24% accuracy, 96.3% sensitivity, and 96.3% F1 Score.
• Models demonstrated low computational cost, suitable for real-time applications, with a DOR of 364.

Conclusions:

• PSO-optimized hybrid models offer accurate and efficient cardiac arrhythmia classification.
• The proposed approach represents a significant advancement in diagnostic performance for clinical decision-making.
• Future research should explore model application to other clinical problems and enhance interpretability.