Paper Title
EFFECTIVE AND ACCURATE HEART DISEASE PREDICTION USING HYBRID MACHINE LEARNING TECHNIQUES BASED ON VALID MAXIMUM ATTRIBUTES

Abstract
This research work presents a comprehensive survey and critical analysis of state-of-the-art machine learning (ML) and hybrid techniques for heart disease prediction, with a focus on accuracy, interpretability, and clinical applicability. The study systematically examines methodologies, datasets (including the widely used UCI Heart Disease dataset), and performance metrics across recent literature (2020–2025), identifying key trends, challenges, and innovations in predictive modeling for cardiovascular diagnostics. The survey highlights the growing adoption of hybrid ML models (e.g., ensemble learning, XGB, deep learning hybrids, and quantum ML) that combine multiple algorithms to improve predictive performance. Techniques such as feature selection (PCA, mRMR, SHAP) and explainable AI (LIME, SHAP) are critically evaluated for their role in enhancing model transparency—a crucial factor for clinical adoption. Furthermore, the study assesses computational efficiency, dataset biases, and real-world deployment challenges, offering insights into scalability and generalizability across diverse populations. Key findings reveal that while hybrid models achieve high accuracy (up to 98%), gaps persist in real-time implementation, dynamic feature engineering, and ethical AI considerations (e.g., data privacy via federated learning). The survey concludes with future research directions, advocating for robust validation frameworks, multimodal data fusion, and clinician-AI collaboration to bridge the gap between theoretical models and practical healthcare applications. This work serves as a foundational reference for researchers and practitioners aiming to advance ML-driven cardiac diagnostics while addressing clinical trust and operational constraints. Keywords - Attributes, Machine Learning, XGBoost, Prediction