Detecting Web Attacks Using Deep Learning

Abstract

Web attack is a cyber-attack which is used to steal the sensitive
information and Un authorization of the computer system without knowing
oneself. Most of the cyber attackers' targets web applications because they
can be accessed easily. When attack is happened an intrusion detection
system sends the alert message to the users and detected. The
implementation of existing methods can be time consuming and more
expensive and requires a lot of domain knowledge. By our prevention
method we use RSMT model to monitor the runtime behaviour of web
application. Then, RSM tool trains an auto encoder which is evoked to
encode the unlabelled data which is used to identify anomalies. After these
results of both datasets application is analysed. Our proposed system can
detect the cyber-attack with small amount of labelled training data