9 Fault Tolerance-Based Attack Detection Using Ensemble Classifier Machine Learning with IOT Security

Computer vision has recently progressed. In our homes and workplaces everywhere, Internet of Things (IoT) devices and applications are being deployed. Deep learning architectures are often fed by continuous data or categorical data collection in these devices. However, since the service provider may make unwanted inferences on the available data, this method poses some privacy and efficiency issues. For simple tasks and lighter models, recent developments in edge processing have paved the way for more effective and private data processing at the source. For larger and more complex versions, however, they remain a challenge. The proposed model is designed between cloud architectures in this paper. The input dataset will be a medical database or sensational dataset; the data has been obtained from IoT. Before this IoT, the cloud structure has been designed with security, and the output of deep learning architecture with ensemble classifiers will have the Intrusion Detection System (IDS) for security. The threshold limit has been set. Similarly, those data beyond the threshold limit will be sensational. Both the cloud architecture will have a similar private key, whether the data is securely transmitted or not. Each information will have its threshold limit, and this has been analyzed; categorized and classified output is obtained from deep learning ensemble classifiers. The experimental results show effective data transmission with higher accuracy when compared with existing techniques.