This repository contains the code and report for the HEP Minor Module Coursework (A3).
- A2: Weight sharing layers for rotation-equivariant image processing
- B1: Higgs boson discovery analysis using ATLAS Open Data
- C2: Event classification for Z boson decay identification using neural networks
.
├── A2_weight_sharing_layers/ # Section A implementation
│ ├── A2_weight_sharing.ipynb # Main notebook
│ ├── figures/ # Generated figures
│ ├── src/ # Source code
│ └── requirements.txt # Dependencies
├── B1_Higgs/ # Section B implementation
│ ├── datasets/ # Data storage
│ ├── notebooks/ # Analysis notebooks
│ ├── src/ # Source code
│ └── requirements.txt # Dependencies
├── C2_event_classification/ # Section C implementation
│ ├── datasets/ # Data storage
│ ├── notebooks/ # Analysis notebooks
│ ├── run/ # Scripts to run experiments
│ ├── src/ # Source code
│ └── requirements.txt # Dependencies
├── report/ # PDF report
└── README.md # This file
- Python 3.8+
- Additional packages as listed in each section's requirements.txt file
# Create a virtual environment (optional)
python -m venv hep_env
source hep_env/bin/activate # hep_env\Scripts\activate on Windows
# Install dependencies for Section A2
pip install -r A2_weight_sharing_layers/requirements.txt
# Install dependencies for Section B1
pip install -r B1_Higgs/requirements.txt
# Install dependencies for Section C2
pip install -r C2_event_classification/requirements.txtThe final report is available in the report/ directory.
- Utilised “Deep research” (OpenAI) and “DeepSearch” (xAI) for high-level overviews and literature searches
- Initial code scaffolding and project structure design
- Template generation for configuration files (
requirements,.gitignore, etc.) - Assistance with network architecture design
- Script prototyping for model training (
.py,.sh) - Debugging support and performance optimisation
- README generation and documentation structuring
- Code commenting and function-level documentation
- Refactoring suggestions to enhance readability
- Implementation optimisation
AI tools used: Claude (Anthropic), ChatGPT (OpenAI), Grok (xAI)