Deep Learning Approaches for Detecting Anomalies and Attacks in IoT Networks

Internet of Things (IoT) is an open and exhaustive system of astute items that have the ability to auto-arrange; assets, information and share data; responding and acting even with circumstance and changes in environment. IoT allows communication of humans-humans, things-things and humans-things. As the volume of devices connected in IoT increases, cybersecurity about and various other types of attacks comes into play. In this paper various cybersecurity attacks like data probing, spying, wrong setup, denial of service, malitious operation, scan and malitious control is are predicted using Deep Learning Model. To predict these attacks a naïve Keras Baseline Model is proposed in this article. Various performance parameters such as Recall, Precision and F1 score is calculated in the classification report and the confusion matrices also generated for evaluating the actual and desired result. The result of our proposed approach is compared with traditional Machine Learning Classifiers such as Logistic Regression and Support Vector Machine. The evaluation matrices shows that the proposed model achieves <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 9. 4 2 \%}$</tex> accuracy.