SQL Injection Attack Detection using Machine Learning Algorithm
A. Sivasangari, Jyotsna Jyotsna, K. Pravalika
Abstract
Injection attack refers to the insertion of malicious code within the network that fetches all the data from the info to the assailant. This kind of attack is taken into account as currently there is a major downside in internet security. The detection of a SQL injection attack continues to be a tough downside because the admin of the info won't apprehend the attack that is going on till there's amendment within the content of the info. The way to successfully sight the injection attack is the foremost vital part of web applications. Hence, the paper introduces adaboost formula that sight numerous styles of injection attacks. The worth of the each weak tree is given highest weight to get the strong model by updating the weights every step via training the dataset, by adding the input of every layer by calculating average of previous outputs. Results indicate that, the proposed algorithm and program has accurately detected the injection attack effectively than that of the initial options of neural techniques area unit, which is degenerated due to the increasing variety of intermediate layers present within the program.