[Retracted] Smart Heart Disease Prediction System with IoT and Fog Computing Sectors Enabled by Cascaded Deep Learning Model

Table 1

Advantages and problems of traditional methods of IoT healthcare for the heart diseases.

Author [citation]	Methodology	Features	Challenges
[18]	Ensemble learning	(i) The prediction outcomes are offered in real time.; (ii) The high accuracies are attained using very low latencies.	(i) It does not permit cost-optimal execution with distinct fog-cloud cost models and QoS characteristics.
[19]	DLMNN	(i) High security is returned in less time.; (ii) The HD of the patient is recognized in a much appropriate manner.	(i) The performance is not enhanced with various optimization as well as feature selection techniques.
[20]	MSSO-ANFIS	(i) The highest fitness value is attained for the entire iterations.; (ii) It is realized by the available products and wearable technologies in the market.	(i) It is not performed with various optimization and feature selection approaches for enhancing the predictive classifier efficiency.
[21]	MDCNN	(i) It offers a high level of accuracy.; (ii) The prediction and the monitoring systems save the lives through instant intervention.	(i) It is not trained and tested using the fully wearable devices.
[22]	EDCNN	(i) It permits the highly reliable and precise heart disease diagnoses.; (ii) It minimizes the misdiagnoses count harming the patients.	(i) The precision is not enhanced by incorporating the advance artificial intelligence.
[23]	HOBDBNN	(i) It returns high recognition accuracy.; (ii) The abnormal heart patterns are recognized in less time.	(i) The IoT-oriented medical disease diagnostic process is not enhanced by the optimized approaches.
[24]	ANFIS	(i) The alerts are generated in minimum time.; (ii) The utility is also improved by less error rates.	(i) It is not made as a user friendly one.
[25]	SVM	(i) It needs less operational and maintenance cost.	(i) It does not deploy various intelligent models for enhancing the accuracy.