Material for the AI-INFN Advanced Hackathon

AI-INFN, as evolution of the ML-INFN initiative, collects and coordinate the efforts on the development and deployment of Artificial Intelligence technologies relevant to INFN research. As part of our programme, we organize training events to discuss base and advanced Machine Learning topics with time to go through the code. We call them hackathons.

The first, second, and fourth ML-INFN hackathons were targeting Machine Learning beginners. The material used for those events is available in another GitHub repository.

The third and fifth ML-INFN hackathons were targeting advanced users, as well as this first AI-INFN hackathon and the related materials are collected in this repository.

Structure of the repository

Contents is organized per topic in different folders. When documentation beyond the Jupyter notebook is needed, a README.md file is included in the sub-directory.

Contents

• ex: material for the hackathon exercises
  • lhcf-cnn: Use of a multidimensional CNN for particle identification in the LHCf experiment
  • gan-detector: Generative Adversarial Networks as a tool to unfold detector effects
  • asd-diagnosis: Autism Spectrum Disorders (ASD) diagnosis using structural and functional Magnetic Resonance Imaging and Radiomics
  • quantum-ml: Quantum Machine Learning applications: classification, anomaly detection and QUBO problems

Automated testing

Tests on the notebooks are run frequently on the different setups being prepared for the hackathon event.

Run all tests with:

python3 -m pytest tests/test_notebooks.py -v --durations=0

Latest results

CNAF - A100 with MIG (2024-11-24)

85.10s call     tests/test_notebooks.py::test_ex_asd_diagnosis[ai4ni-sMRI_fMRI_sep]
78.64s call     tests/test_notebooks.py::test_ex_asd_diagnosis[ai4ni-Joint_Fusion]
74.39s call     tests/test_notebooks.py::test_ex_quantum_ml[qml-QClassifier_*]
74.38s call     tests/test_notebooks.py::test_ex_gan_detector[gan-k2]
59.21s call     tests/test_notebooks.py::test_ex_lhcf_cnn[cnn-k3]
48.38s call     tests/test_notebooks.py::test_ex_lhcf_cnn[cnn-k2]
28.07s call     tests/test_notebooks.py::test_ex_quantum_ml[qml-QAE_*]
12.98s call     tests/test_notebooks.py::test_ex_quantum_ml[qml-QUBO_*]
10.03s call     tests/test_notebooks.py::test_env_tensorflow[gan-k3]
10.03s call     tests/test_notebooks.py::test_env_quantum[qml]
9.00s call     tests/test_notebooks.py::test_env_tensorflow[cnn-k2]
8.53s call     tests/test_notebooks.py::test_env_tensorflow[cnn-k3]
7.67s call     tests/test_notebooks.py::test_env_tensorflow[gan-k2]
7.56s call     tests/test_notebooks.py::test_env_tensorflow[ai4ni]

ReCaS - A100 with MIG (2024-11-24)

86.77s call     tests/test_notebooks.py::test_ex_quantum_ml[qml-QClassifier_*]
80.08s call     tests/test_notebooks.py::test_ex_asd_diagnosis[ai4ni-Joint_Fusion]
72.02s call     tests/test_notebooks.py::test_ex_asd_diagnosis[ai4ni-sMRI_fMRI_sep]
64.25s call     tests/test_notebooks.py::test_ex_gan_detector[gan-k2]
56.46s call     tests/test_notebooks.py::test_ex_lhcf_cnn[cnn-k3]
50.56s call     tests/test_notebooks.py::test_ex_lhcf_cnn[cnn-k2]
29.88s call     tests/test_notebooks.py::test_ex_quantum_ml[qml-QAE_*]
12.94s call     tests/test_notebooks.py::test_ex_quantum_ml[qml-QUBO_*]
7.60s call     tests/test_notebooks.py::test_env_quantum[qml]
6.37s call     tests/test_notebooks.py::test_env_tensorflow[gan-k3]
6.34s call     tests/test_notebooks.py::test_env_tensorflow[cnn-k2]
6.27s call     tests/test_notebooks.py::test_env_tensorflow[cnn-k3]
6.11s call     tests/test_notebooks.py::test_env_tensorflow[ai4ni]
5.99s call     tests/test_notebooks.py::test_env_tensorflow[gan-k2]

License

Code is released under OSI-approved MIT license.

The documentation provided in the form of Jupyter notebooks is released under CC-BY-NC-SA license.