Detection and Classification of DDoS Attacks in Cloud Data Using Hybrid LSTM and RNN for Feature Selection

Cyber threats are the main source of concern in the sector it is deployed. DDoS attacks are at the very top of the list. The rapid growth of cloud migration also expands the attack surface. Different forms of DDoS attacks exist, such as DoS and DDoS etc., There are several existing machine learning algorithms to detect DDoS attacks. There are also many hybrids of machine learning algorithms which are introduced for detecting DDoS attacks to increase the accurateness of detection. Here we use hybrid Long Short-Term Memory Network (LSTM) and Recurrent Neural Network (RNN) for feature selection and deliver an improved detection system. The proposed exploratory model of Multi-layer Perceptron (MLP) classifiers uses deep neural network algorithms. The final outcome shows higher accuracy and precision by dealing with a huge dataset that was gathered. The accuracy and precision of 98.85% and 92 % respectively for grouping in structure.