The Internet of Things application is currently confronted with novel ultimatum as the dimension of sensitive data grows. To secure data and ensure privacy, a judicious IoT system is required. In recent days machine learning has offered a propitious solution for securing sensitive data and devices of Interne of Things application. This paper proposed a framework for securing Internet of Things application. The proposed scheme is mainly developed for real time data monitoring with preserving confidentiality and classification of normal and abnormal traffic. A cluster-based approach has been considered here for decision purposes and real time analysis for securing application. The numerical result of proposed framework outperforms in respect of benchmark models in terms of accuracy, false positive rate and area under the cover