Application of Logistic Regression in WEB Vulnerability Scanning
Tao Liu, Longtao Zhang · 2018
In the anomaly detection case of the web site, the traditional vulnerability scanning technology is com-monly used for active defense at present, but there are some problems such as incomplete data crawling and time consuming iteration of rule set, Therefore, this paper puts forward a logistic regression model based on machine learning to detect anomaly of web traffic, using TF-IDF algorithm to complete the feature extraction of URL, then training the logistic regression model, it is found that using LBFGS algorithm can be used as a loss function to obtain better model accuracy. At the same time, the experimental results show that the model can distinguish abnormal traffic well for newly generated web traffic.