MA-LSTMNet: AN EFFICIENT MULTISCALE ATTENTION LSTMNet-BASED HEART DISEASE PREDICTION FRAMEWORK WITH IoT DEVICES INCORPORATING HYBRID RAT SWARM-RED DEER ALGORITHM

Worldwide, Heart Disease (HD) has become one of the deadly diseases generating a high mortality rate. Different techniques are adapted to analyze HD which are highly complicated. Furthermore, machine learning techniques are highly performed to offer a better performance rate in identifying and diagnosing cardiac diseases. However, these techniques couldn’t handle the high-dimensional data. This research work is focused to develop an efficient IoT-based HD monitoring framework using the multiscale attention-based deep learning architecture. Initially, the IoT-based text and signal data are gathered from online sources. Then, the pre-processed data are used for the deep feature retrieval phase and signal feature extraction. Here, the deep feature extraction is performed using Deep Convolutional Neural Network (DCNN) and signal features. Both deep features and signal features are provided to the weighted feature fusion region, where the weights are modified, and optimal features are selected using the Hybrid Rat Swarm with Red Deer Optimization Algorithm (HRS-RDOA) for enhancing the prediction performance. Then, the weighted fused features are used in the HD forecast region, where the Multiscale Attention-based Long Short-term Memory Network (MA-LSTMNet) is employed to predict heart disease. Here, the parameter optimization takes place using developed HRS-RDOA for getting accurate predicting results. The experimental analyses contrast the implemented IoT-based HD prediction model and conventional methods. While validating through various performance measures, the accuracy of the developed model is 96% and also the precision is 95%. Throughout the validation, the offered model shows enhanced performance than the existing methods. Generally, healthcare applications are considered for predicting HD where the clinicians can treat the HD in an effective manner which leads to reduction in the mortality rate.