Privacy-Preserving Analytics of IoT Data Using Generative Models

The Internet of Things (IoT) is rapidly growing and affecting various aspects of our daily lives, allowing for data collection and sharing. The number of devices connected to the IoT environment is growing and implies instant surveillance of almost every action, sound, and move we encounter daily. This vast spread and usage of wearable devices raises the issue of high possibilities of several types of security concerns. The wide growth of such devices implies a substantial risk in data generation, storage, and transmission via IoT devices. Numerous studies have proposed privacy-preserving methods aiming to provide better privacy. However, the exponential growth of IoT data has resulted in a decline in the efficiency and performance of existing methods. This chapter proposes a generative privacy-preserving model for IoT data. The work in this research proposes a generative model to perturb data while preserving its features. The performance of the proposed model is measured using several criteria, such as classification accuracy, precision, recall, the area under the curve (AUC), and F-score. The evaluation metrics used show remarkable results in terms of accuracy and privacy.