Anomaly Detection Visualization

Ready-to-use Jupyter notebooks for generating presentation-quality plots demonstrating anomaly detection methods (kNN, Isolation Forest, PCA, HBOS).
Used in the Positive Hack Days conference talk. Fork and run to reproduce slides or adapt for your needs.

Features

• Plug & Play: Pre-configured notebooks with example datasets.
• Conference-Ready: Clean, styled visuals (Seaborn/Matplotlib).
• Algorithms Covered:
  • k-Nearest Neighbors (kNN)
  • Isolation Forest
  • Principal Component Analysis (PCA)
  • Histogram-Based Outlier Detection (HBOS)

Quick Start

1. Clone the repo:

   git clone https://github.com/onixlas/phd-2025-anomaly-detection.git

2. Install dependencies:

    pip install -r requirements.txt

3. Run any notebook.

Tech Stack

• Python + Jupyter Notebook
• Libraries: Scikit-learn, Pandas, NumPy, PyOD, Seaborn