Smart Intelligence Based Cardio-Heart Disease Classification using Deep Feature Engineering with Adaptive Fuzzy Integrated Multilayer Perceptron Neural Network

heart disease is very dangerous to causes series death that affects suddenly. The indications may be silent or it shows some signs at a preliminary stage. The healthcare sector is fast growing to identify the disease feature signs earlier to predict and categorize the type of disease to make early diagnosis and treatment. The development of Artificial intelligence and deep feature engineering models are used to analyze the medical features to find the disease. The traditional systems are complex structures and feature selection leads to nonrelational precision accuracy due to feature dimension which leads to poor classification accuracy. To resolve this problem, to intend a new Adaptive Fuzzy Integrated Multilayer Perceptron Neural Network (AFMPNNet) with Mean Absolute Coefficient Feature Selection (MACFS) to classify the heart disease effectively. The dataset is collected from healthcare cardio-Ds for effective preprocessing using the Z-Score Scaling (ZSS) method to verify the actual ideal limits of feature labels. Then Mean Absolute Coefficient Feature Selection (MACFS) is applied to extract the features in which the ideal margin mean rate is crossed over the mutual limits in muti-feature cross limits at exactly the disease-affected features. The feature selection reduces the non-relation feature dimension to improve the sensitivity accuracy. Finally, the fuzzified unit is designed at the dense unit to MPNNet to adaptive active unit for enhancing the classification accuracy which predicts the disease classes. The proposed system produces high performance compared to other systems as well in heart disease prediction accuracy presents improved sensitivity, specificity, and F1 measure with redundant false rate.