Threat Detector using Machine Learning Rules

The Art and Science of Cybersecurity Attack Detection:A Hibrid Approach

Developed by : Lavinia-Cristiana Bacaru; 💻: @codinglavinia

---

🇪🇸 Descripción del proyecto:

Sistema de Detección de Amenazas tipo I.D.S basado en reglas de Machine Learning para identificar tráfico de red malicioso utilizando el dataset UNSW-NB15.

Mi proyecto aplica técnicas de preprocesamiento, análisis exploratorio y clasificación supervisada para detectar múltiples tipos de ciberataques con alta precisión.

---

🇬🇧 Project description:

Machine Learning-based Intrusion Detection System (IDS) designed to identify malicious network traffic using the UNSW-NB15 dataset.

My project applies preprocessing, exploratory analysis and supervised classification techniques to detect multiple cyberattack categories with strong performance.

---

⚙️ Python Libraries for Data Science in this project :

---

Methodology / Metodología:

• Data cleaning & preprocessing
• Feature engineering
• Supervised ML classification
• Model evaluation (Accuracy, Confusion Matrix, F1-score)

🤖🔐My Python Project Architecture:

          +-----------+
          | Raw Data  |
          |  Step 1   |
          +-----------+
                 │
                 ▼
      +-------------------+
      | Pandas: Cleaning  |
      |      Step 2       |
      +-------------------+
                 │
                 ▼
      +--------------------+
      | NumPy: Feature Eng |
      |       Step 3       |
      +--------------------+
                 │
                 ▼
      +--------------------+
      | Scikit-learn: ML   |
      | Training           |
      |      Step 4        |
      +--------------------+
                 │
                 ▼
      +--------------------+
      | Model Eval & Metrics|
      |       Step 5        |
      +--------------------+
                 │
                 ▼
      +----------------------------+
      | Matplotlib/Seaborn:        |
      | Visualization & Insights   |
      |           Step 6           |
      +----------------------------+
                 │
                 ▼
          +-----------+
          | Results   |
          |  Step 7   |
          +-----------+

---

📂 Dataset :

🔗 https://www.kaggle.com/datasets/dhoogla/unswnb15

Required files: 1 .UNSW_NB15_training-set.csv and 2. UNSW_NB15_testing-set.csv;

---

▶️ Run

git clone https://github.com/tuusuario/turepositorio.git
cd turepositorio
pip install pandas numpy matplotlib seaborn scikit-learn
jupyter notebook