Enhanced Anomaly Detection in Keystroke Dynamics Authentication

This repository is part of the CS266 - Topics in Information Security course project titled "Enhanced Anomaly Detection in Keystroke Dynamics Authentication". The project was submitted as part of the course requirements by the team members:

• Rashmi Sonth
• Sakshi Sanskruti Tripathy

The project focuses on analyzing different methodologies for anomaly detection in keystroke dynamics authentication, leveraging advanced machine learning and deep learning models.

---

File Structure

1. Dataset

• demographics.csv:

  • Contains demographic information of participants.
  • Used to supplement analysis with demographic-based features.

• free-text.csv:

  • Contains raw keystroke dynamics data.
  • Includes timing features such as DU.key1.key1, DD.key1.key2, etc.
  • This dataset is critical for building and testing machine learning models.

• Dataset Source:

  • The dataset used in this project can be downloaded from Zenodo.

2. Notebook

• models.ipynb:
  • The main Jupyter Notebook for training and evaluating models.
  • Implements machine learning and deep learning approaches (e.g., LSTM, CNN).
  • Important: Ensure the dataset paths (free-text.csv and demographics.csv) are correctly updated in this notebook based on your local environment.

3. Output

• output.png:
  • Helps to evaluate the effectiveness of implemented techniques.

---

Instructions

1. Update Dataset Paths:

   • The datasets are located in the dataset folder.
   • Ensure you update the dataset paths in the models.ipynb file wherever necessary:

     data = pd.read_csv('dataset/free-text.csv')
     demographics = pd.read_csv('dataset/demographics.csv')

2. Run the Notebook:

   • Open models.ipynb in Jupyter Notebook or any compatible environment.
   • Execute the cells step-by-step to train and evaluate the models.

---

Requirements

• Python 3.7+

• Libraries:

  • tensorflow
  • numpy
  • pandas
  • matplotlib
  • seaborn
  • sklearn

  ---

  Quick Local Run

  1. Install dependencies:

  python -m pip install -r requirements.txt

  2. Run the main script (uses dataset/free-text.csv and dataset/demographics.csv):

  python models.py

  Note: the script will automatically fall back to the local dataset/ folder when not running in Google Colab.

  ---

  Full Setup & Run (Windows)

  Option A — Recommended (conda, fastest & avoids build issues)

  1. Create and activate an environment with conda:

  conda create -n keystroke python=3.11 -y
  conda activate keystroke

  2. Install binary packages from conda-forge:

  conda install -c conda-forge pandas numpy scipy scikit-learn matplotlib seaborn -y

  3. Run the script:

  python models.py

  Option B — Virtualenv (works but may require build toolchain)

  1. Create the venv (PowerShell):

  python -m venv .venv

  2. Use the venv's Python directly (no activation required):

  .venv\bin\python.exe -m pip install --upgrade pip setuptools wheel
  .venv\bin\python.exe -m pip install -r requirements.txt
  .venv\bin\python.exe models.py

  Note: on Windows some venvs use Scripts instead of bin (check .venv for either bin or Scripts).

  Troubleshooting

  • SSL / certificate verify failures while pip installing: This environment shows ssl.SSLCertVerificationError when pip attempts to download build backends (cmake, ninja). Common fixes:

    • Use the Conda approach above (conda provides prebuilt binaries). This is recommended on Windows.
    • Ensure system time is correct and corporate proxies are configured. If your network intercepts TLS, install the organization's CA into the OS certificate store.
    • As a last resort, set the environment variable PIP_DISABLE_PIP_VERSION_CHECK=1 and use --trusted-host for pip (not recommended for long-term use).

  • Build errors for numpy/pandas: Building numpy and pandas from source on Windows often requires a C/C++ toolchain (MSVC) and CMake. Use conda to avoid this.