Botnet Attack Detection in Machine Learning Using ONE API
- 0 Collaborators
Botnet attacks are the most serious threat to security attack . To detect it we have to detect unusual traffic, identifying suspicious devices and IP addresses, and eliminating communication with suspicious actors. ...learn more
Project status: Published/In Market
Intel Technologies
DevCloud
Overview / Usage
The small program to perform any type of malicious activity that may damage the system of the legal user automatically without legal users' knowledge is called a bot (bad bot). The network of bots under the control of a botmaster is called a botnet. It is a serious threat to information, communication, and economy etc. The interaction of devices to form a botnet are smartphones, computers, IoT systems whose vulnerabilities are exploited and so the security is breached to relinquish the control to bot controllers or third-party.
Methodology / Approach
1.Import libraries
2.Understand the features and visualise it
3.Replace null with Zero.
4.The features such as 'src_ip', 'conn_state', 'src_port', 'dst_ip', 'dst_port', and 'ts
5.We converted the categorical features into numerical using the LabelEncoder() function from the sklearn library of python. We used label encoding despite the one-hot encoding because one-hot encoding increases the dimensionality of the dataset by adding an extra column of every single category.
6.Test with different classification algorithms and find the best
7.Train the model in One Api for the better results with fast computation
8.Test the accuracy with the algorithms and find the best
Technologies Used
numpy
pandas
matplotlib
One DAL of one API