LSTM Based Multiple Squashing Functions Deep Learning Model for Advanced Traffic Management System Attack Detection
Chethana Dukkipati, Yunpeng Zhang, Liang Chieh Cheng · 2018
Advanced Traffic Management System is started as a standalone system and now with the developed technology, it has turned to more complex networks through which many types of attacks are occurring on the system. These attacks are difficult to identify and to resolve them. In this paper, we are focusing on detecting these attacks present in the Advanced Traffic Management systems through the booming concept “Deep learning model”. The model proposed in the paper is the extension of Long Short-Term Memory(LSTM) recurrent model. The basics of Recurrent Neural Networks(RNN) model is taken due to its feedback behavior. The model is applied on a dataset that is collected from the online website and validated the model by coding it in python. By the end of the code execution, the model gives us whether the given test data has any attack in it or not. The model checks for more number of features for attack detection than the LSTM.