Modelling a Dense N Model for Anomaly Prediction in IoT Environment
Mohan Kumar Chandol, M. Kameswara Rao, Chandra Sekhar Kolli · 2022 4th International Conference on Inventive Research in Computing Applications (ICIRCA) · 2022
IoT (Internet of Things) devices frequently disclose sensitive data explicitly or implicitly. They may be able to detect whether the user is currently at home are not. As a result, it is critical to safeguard those devices from attackers and trigger prior warning who alerts compromised equipment in a timely manner. This research presents temporal window embedding techniques that efficiently process vast amounts of data with minimal memory. The proposed study employs an anomaly detection unit and the recommended embedding vectors based on the encoder elements of the transformer, accompanied by feed-forward network. The anticipated model is compared with the existing strategies of conventional learning approaches. As a result, several learning techniques are explicitly analyzed and the efficacy of IoT has been examined. The proposed hypothesis is validated by extensive experimentation on the new IoT-23 dataset.