Advanced privacy protection (APP) machine learning model using cryptographic techniques for IoT
K Senthil, R. Karthikeyan, S. Shanmuga Priya, R Monikaa, S. Ramamoorthi, S. Faizal Mukthar Hussain
Abstract
Because of the continuous computational and communication growth, the Internet of Things (IoT) plays a significant role in many real-time applications. Hence, huge amount of data are produced by the IoT devices, requires privacy preserving models for securing the data. For preserving the privacy of data, many machine learning models are developed, still, certain models lack in efficiency. For this, an Advanced Privacy Protection (APP) Machine Learning Model is proposed in this paper. The model uses cryptographic techniques for preserving the data privacy in efficient manner. Moreover, the model contains a Secure Data Provider (SDP) for processing the privacy protection-based training with the data on the nodes. The data privacy is ensured and the model factors can be acquired by SDP, where Support Vector Machine (SVM) is the training model employed. The results show that the model significantly increases the accuracy of training model and data privacy, minimizes the communication overhead, computational complexities.