Optimized Layered Sparse Representation Network (OLSRN) for Accurate Cardiovascular Disease Prediction

Abstract

: Heart disease remains a leading cause of mortality worldwide, claiming millions of lives each year and posing a significant challenge to global healthcare systems. The early detection and accurate diagnosis of cardiovascular conditions are critical in addressing this pressing issue. By identifying potential risks and enabling timely medical intervention, these measures play a pivotal role in improving patient outcomes and significantly reducing the associated mortality rates. While state-of-the-art machine learning algorithms, such as RF, SVM, and XGBoost, have demonstrated promising results, they are often hindered by limitations in feature extraction, parameter optimization, and adaptability to complex data. To overcome these challenges, this research introduces the Optimized Layered Sparse Representation Network (OLSRN), a novel framework tailored for precise CVD prediction. OLSRN integrates an encoder-decoder mechanism for effective feature extraction, leveraging sparse autoencoders to capture intricate data patterns. A softmax classifier is incorporated to enhance classification performance, while a modified PSO algorithm is employed to fine-tune model parameters, ensuring optimal accuracy and efficiency. By limiting the network depth to three sparse autoencoders, OLSRN strikes a balance between computational complexity and performance. Extensive evaluations on UCI benchmark datasets demonstrate the superiority of OLSRN over conventional methods, with significant improvements in accuracy, precision, recall, and F1-score. This research highlights the potential of OLSRN as a robust and scalable solution for CVD prediction, paving the way for advancements in healthcare diagnostics and machine learning applications.