HEPML Resources

Listing of useful (mostly) public learning resources for machine learning applications in high energy physics (HEPML). Listings will be in reverse chronological order (like a CV).

N.B.: This listing will almost certainly be biased towards work done by ATLAS scientists, as the maintainer is a member of ATLAS and so sees ATLAS work the most. However, this is not the desired case and help to diversify this listing would be greatly appreciated.

Table of contents

• Introductory Material
  • Lectures
  • Seminar Series
  • Tutorials
  • Schools
  • Courses
  • Journals
• Software
• Public Datasets
• Papers
• Workshops
• Tweets
• Other HEP Resource Collections
• People
• Contributing

Introductory Material

Lectures

• Introduction to GANs, by Luke de Oliveira (November 3, 2017)

• Frontiers with GANs, by Michela Paganini (November 3, 2017)

• Nikhef Colloquium: "Teaching machines to discover particles", by Gilles Louppe (September 29, 2017)

• CERN Academic Training Lecture Regular Programme, April 2017 (Machine Learning):

  • Machine Learning (Lecture 1) --- Michael Kagan (SLAC)
  • Machine Learning (Lecture 2) --- Michael Kagan (SLAC)
  • Deep Learning and Vision --- Jonathon Shlens (Google Research)

• Deep Learning in High Energy Physics, by Amir Farbin

Seminar Series

• CERN Data Science Seminars

• Inter-Experimental LHC Machine Learning Working Group Guest Seminars:

  • Open challenges for improving Generative Adversarial Networks (GANs), by Ian Goodfellow (October 27, 2017)

Tutorials

• PyTorch Deep Learning Minicourse - CoDaS-HEP 2018, by Alfredo Canziani

• Intro Tutorial on GANs, by Michela Paganini

• Scikit-learn Tutorial, by Gilles Louppe

• TMVA Tutorial, by Lorenzo Moneta

• Keras and TMVA interfaces Tutorial, by Stefan Wunsch

• Boosted Decision Tree Tutorial (using XGBoost), by Katherine Woodruff

• Introduction to Deep Learning with Keras Tutorial, by Luke de Oliveira

• Introduction to Deep Learning with Keras Tutorial - 2nd Developers@CERN Forum, by Michela Paganini

Schools

HEP-ML:

Upcoming:

• Deep Learning for Science Summer School 2019, Berkeley, CA, USA (July 15-19, 2019)

Past:

• 5th Machine Learning in High Energy Physics Summer School 2019 (July 1-10, 2019)
  • Associated Yandex School of Data Analysis repo: mlhep2019
• 4th Machine Learning in High Energy Physics Summer School 2018 (August 6-12, 2018)
  • Associated Yandex School of Data Analysis repo: mlhep2018
• 2nd Computational and Data Science school for High Energy Physics (CoDaS-HEP 2018) (July 23-27, 2018)
• 3rd Machine Learning in High Energy Physics Summer School 2017 (July 17-23, 2017)
  • Associated Yandex School of Data Analysis repo: mlhep2017
• 1st Computational and Data Science School for High Energy Physics (CoDaS-HEP) (July 10-13, 2017)
• 2nd Machine Learning in High Energy Physics Summer School 2016 (June 20-26, 2016)
  • Associated Yandex School of Data Analysis repo: mlhep2016
• 1st Machine Learning in High Energy Physics Summer School 2015 (August 27-30, 2015)
  • Associated Yandex School of Data Analysis repo: mlhep2015

Deep Learning:

Upcoming:

• Machine Learning Summer School 2019, London, UK (July 15–26, 2019)
• Machine Learning Summer School 2019, Stellenbosch, South Africa (January 7-18, 2019)

Past:

• Machine Learning Summer School 2018, Madrid, Spain (August 27 - September 7, 2018)
• Machine Learning Summer School 2018, Buenos Aires, Argentina (June 18-20, 2018)
• Deep Learning and Reinforcement Learning Summer School, Toronto, Canada (July 25 - August 3, 2018)
• PAISS: Artificial Intelligence Summer School, Grenoble, France (July 2-6, 2018)
• Deep Learning Summer School 2016 (August 1-7, 2016)

Courses

• Advanced Machine Learning, Pierre Geurts, Gilles Louppe, and Louis Wehenkel (Spring, 2018 - Université de Liège, Institut Montefiore)

• Applications of Deep Learning to High Energy Physics, Amir Farbin (Spring, 2017 - University of Texas at Arlington)

  • Associated GitHub repository: DSatHEP-Tutorial

• Tensorflow for Deep Learning Research, (Spring, 2017 - Stanford Univeristy)

• Introduction to Machine Learning and Convolutional Neural Networks for Visual Recognition:

  • Spring, 2017 - Stanford University, Fei-Fei Li, Justin Johnson, Serena Yeung
  • Winter, 2016 - Stanford University, Andrej Karpathy, Fei-Fei Li, Justin Johnson

Journals

• Distill Research Journal

Software

Common software tools and environments used in HEP for ML

• Python environments for scientific computing

  • The Conda package and environment manager and Anaconda Python library collection

    • Using ROOT/PyROOT with Conda and NumPy

  • scikit-learn: General machine learning Python library

• TMVA: ROOT's builtin machine learning package

  • TMVA-branch-adder: wrapper to add TMVA response to TTree without boiler plate code

High level deep learning libraries/framework APIs

• Keras

Deep learning frameworks

• TensorFlow
• Theano
• PyTorch
• Caffe2
• List of Conversion Tools For Saved Networks

HEP to ML bridge tools

• lwtnn: Tool to run Keras networks in C++ code

• sklearn-porter: Transpile trained scikit-learn estimators to C, Java, JavaScript and others

• ONNX open format to represent deep learning models

• Scikit-HEP: Toolset of interfaces and Python tools for Particle Physics

  • root_numpy: The interface between ROOT and numpy

  • root_pandas: An upgrade of root_numpy to use with pandas

  • uproot: Mimimalist ROOT to numpy converter (no dependency on ROOT)

• ttree2hdf5: Mimimalist ROOT to HDF5 converter (written in C++)

• hep_ml: Python algorithms and tools for HEP ML use cases

Images for Containerized Environments

• ATLAS Machine Learning Docker images: Base images for a modern Python 3 machine learning environment for physics

Public Datasets

• CERN IML public datasets listing: Listing of public datsets that are used for machine learning studies at the LHC.

Papers

A .bib file for all papers listed is available in the tex directory.

A listing of papers of applications of machine learning to high energy physics can be found in papers.md.

Workshops

Upcoming

• TBA

Past

• Machine Learning for Jet Physics (2020) (January 15-17, 2020)
• Machine Learning and the Physical Sciences at NIPS (December 14, 2019)
• 4th ATLAS Machine Learning Workshop (2019) (November 11-15, 2019)
• Fast Machine Learning IRIS-HEP Blueprint Workshop (September 10-13, 2019)
• 3rd CMS Machine Learning Workshop (2019) (June 17-19, 2019)
• Theoretical Physics for Deep Learning at ICML 2019 (June 14, 2019)
• 3rd IML Machine Learning Workshop (2019) (April 15-18, 2019)
• Accelerating the Search for Dark Matter with Machine Learning (2019) (April 8-12, 2019)
• 5th Connecting The Dots / Intelligent Trackers (2019) (April 2-5, 2019)
• 19th International Workshop on Advanced Computing and Analysis Techniques in Physics Research (ACAT 2019) (March 11-15, 2019)
• Machine Learning for Jet Physics (2018) (November 14-16, 2018)
• 3rd ATLAS Machine Learning Workshop (2018) (October 15-17, 2018)
• 2nd CMS Machine Learning Workshop (2018) (July 2-4, 2018)
• 2nd IML Machine Learning Workshop (2018) (April 9-12, 2018)
• Machine Learning for Phenomenology (2018) (April 3-6, 2018)
• 4th International Connecting The Dots Workshop (2018) (March 20-22, 2018)
• Accelerating the Search for Dark Matter with Machine Learning (2018) (January 15-19, 2018)
• Machine Learning for Jet Physics (2017) (December 11-13, 2017)
• Deep Learning for Physical Sciences at NIPS (December 8, 2017)
• 18th International Workshop on Advanced Computing and Analysis Techniques in Physics Research (ACAT 2017) (August 21-25, 2017)
• Hammers & Nails - Machine Learning & HEP (July 19-28, 2017)
• CMS Machine Learning Workshop (2017) (July 5-6, 2017)
• 2nd ATLAS Machine Learning Workshop (2017) (June 6-9, 2017)
• Workshop on Machine Learning and b-tagging (May 23-26, 2017)
• DS@HEP 2017 (May 8-12, 2017)
• 2nd S2I2 HEP/CS Workshop (Parallel Session) (May 1-3, 2017)
• CERN openlab workshop on Machine Learning and Data Analytics (April 27, 2017)
• First IML Workshop on Machine Learning (March 20-22, 2017)
• DS@HEP at the Simons Foundation (July 5-7, 2016)
• ALICE Mini-Workshop 2016: Statistical Methods and Machine Learning Tutorial (May 18, 2016)
• ATLAS Machine Learning Workshop (2016) (March 29-31, 2016)
• Heavy Flavour Data Mining workshop (February 18-20, 2016)
• Data Science @ LHC 2015 (November 9-13, 2015)

Tweets

• #HEPML collection of tweets

People

• HEPML directory: Opt-in list of people working at the intersection of Machine Learning and High Energy Physics
  • Add yourself through the Google form

Other HEP Resource Collections

• HEP Software Foundation's list of Python Libraries of Interest to Particle Physics

Contributing

Contributions to help improve the listing are very much welcome! Please read CONTRIBUTING.md for details on the process for submitting pull requests or filing issues.

Authors

Listing maintainer: Matthew Feickert

Acknowledgments

• Following PurpleBooth's README style
• All badges made by shields.io
• Inspiration for this listing came from the Awesome Machine Learning repo and Dustin Tran's Machine Learning Videos repo
• Many thanks to everyone who has contributed their time to improve this project