Botnet Detection using Machine Learning

A machine learning project to detect botnet-infected network traffic.

Trained on the CTU-13 dataset and built using Python.

---

Features and Project Description

• Clean & preprocess CTU-13 botnet data
• Train a Random Forest model
• Predict botnet traffic from test files
• Interactive Streamlit dashboard

---

Setup

1. Clone the repo

    git clone https://github.com/Krishna9588/Botnet-Detection-using-ML.git

2 Install dependencies

    pip install -r requirements.txt

3 Download Dataset

    git clone https://github.com/imfaisalmalik/CTU13-CSV-Dataset
    cd dataset

---

About Dataset and dataset folder

In this project I have used CTU-13 Botnet Dataset. Which is available on Git Hub for public use.

1. CTU13_Attack_Traffic - Dataset with Bot Attack
2. CTU13_Normal_Traffic - Normal User Dataset
3. prediction_output - Result dataset
4. test_sample - Sample dataset use for getting output

---

Run this Project

1. Preprocess & Training the model

    python train_model.py

2. Detection

    python detector.py

3. Dashboard

    streamlit run streamlit_app.py

---

Expected Output

SrcAddr, DstAddr, Proto, TotBytes, ..., Prediction, Prediction_Label
192.168.1.10, 8.8.8.8, TCP, 2400, ..., 0, NORMAL
10.0.0.5, 192.168.1.3, TCP, 5500, ..., 1, BOTNET

---

Next thing I am trying in this Project is to use my real-time traffic using the PyShark Python library as input to the model.

---