README.md

http_web_intusion

Predict whether the HTTP request is normal or an anomaly depending on the set of parameters using various Machine Learning Models.

Features taken in consideration for this project

1. HTTP Method
2. Length of URL
3. Length of Argument
4. Number of digits present in the Argument
5. Number of characters present in the Argument
6. Number of special characters present in the Argument
7. Number of digits present in the URL
8. Number of Characters present in the URL
9. Number of Special Characters present in the URL
10. Number of digits present in Cookies
11. Number of Characters present in Cookies
12. Length of Content-Type
13. Length of Content-Length
14. Number of Keywords present in the argument

Machine Learning Algorithms used

1. Logistic Regression (LR)
2. Stochastic Gradient Descent (SGD)
3. Multilayer Perceptron (MLP)
4. Support Vector Machine (SVM)
5. Linear Discriminant Analysis (LDA)
6. CART or Bagging Classifier

Results

Algorithm	Accuracy
Stochastic Gradient Descent	95.86
Multilayer Perceptron	98.74
Support Vector Machine	95.98
Linear Discriminant Analysis	92.34
Bagging Classifier	99.31

Springer Chapter Link:

HTTP Request Anomaly Detection using Machine Learning