Heart Disease Classification using Adaptive Recurrent Neural Network Method

Cardiovascular diseases are still among the leading causes of mortality, which require fast and reliable classification methods for diagnosing and preventing the disease. This work thus presents an elaborate framework involving an ARNN for heart disease categorization. First, the technique of Variational Autoencoders (VAE) is used to detect and remove irrelevant or undesirable data records, purging the dataset. Subsequent to this, the Information Gain Feature Selection (IGFS) technique is employed to select the most informative features that lower dimensionality and increase the model’s accuracy. Finally, the ARNN is used for classification using the fact that it is a dynamic network that can accommodate temporal characteristics of the data. The experimental results also show that the accuracy, precision, and recall of the approach are higher and higher than those of the conventional method. The proposed framework demonstrates tremendous promise for real-world implementation in medical diagnostics and is a sound and scalable approach to heart disease prediction. Even more striking, these results stress the meaning of applying modern techniques to machine learning in healthcare-sector development. The ARNN achieve 95.6% accuracy for predict the cardiovascular disease.