Mumbai-Pune Collider Physics Initiative

Mumbai Pune Collider Meet Fall 2017

15th October 2017
IIT Bombay, India

Meeting agenda

The morning session of the meeting will cover talks by students and postdocs on their latest results in the field of collider physics. If you are interested in giving a talk or having your student give a talk, please email the organizers. A rough sample of topics that people are working on include:

  1. Soft Tracks
  2. QTrimming
  3. W Polarization in hadronic decays
  4. Top polarization
  5. Lepton jets

Any other work on collider physics and new signatures/search strategies is welcome for discussion.

The afternoon session of this meeting will be focused on machine learning. This is an area that is relatively new to most of the participants.

Recently there has been a burgeoning interest in the HEP community towards the adoption of novel machine learning (ML) techniques [1, 2, 3, 4, 5, 6, 7, 8]. Techniques leveraging artificial intelligence methodologies, for instance from from Deep Learning (DL), show promise in tagging objects and in pile-up mitigation. In [1], the basic idea was to use DL to help discriminate gluon and quark jets by treating the calorimetric energy deposits akin to an “image”. [2] introduced the paradigm of classification without labels (CWoLa) where a classifier is trained to distinguish statistical mixtures of HEP relevant classes. The technique minimises the risk of ML algorithms latching on artefacts in MC simulations. There have also been very interesting attempts on top-tagging [3] and pile-up mitigation techniques in this context [4].

As part of our Mumbai-Pune Collider initiative, one possibility would be to have a general discussion on some of the recent ML techniques

  • “Weaky-supervised” strategies
  • Convoluted Neural Networks
  • Deep Learning

This may be followed by a discussion on where our respective groups can focus on, as an application to specific problems. One straightforward speculation would be whether some of the ML methods can be directly applied to the currently ongoing projects.

References

  1. P. T. Komiske, E. M. Metodiev and M. D. Schwartz, “Deep learning in color: towards au- tomated quark/gluon jet discrimination,” JHEP 1701, 110 (2017) doi:10.1007/JHEP01(2017)110 [arXiv:1612.01551 [hep-ph]].
  2. E. M. Metodiev, B. Nachman and J. Thaler, “Classification without labels: Learning from mixed samples in high energy physics,” arXiv:1708.02949 [hep-ph].
  3. A. Butter, G. Kasieczka, T. Plehn and M. Russell, “Deep-learned Top Tagging using Lorentz Invariance and Nothing Else,” arXiv:1707.08966 [hep-ph].
  4. P. T. Komiske, E. M. Metodiev, B. Nachman and M. D. Schwartz, “Pileup Mitigation with Machine Learning (PUMML),” arXiv:1707.08600 [hep-ph].
  5. T. Cohen, M. Freytsis and B. Ostdiek, “(Machine) Learning to Do More with Less,” arXiv:1706.09451 [hep-ph].
  6. J. Pearkes, W. Fedorko, A. Lister and C. Gay, “Jet Constituents for Deep Neural Network Based Top Quark Tagging,” arXiv:1704.02124 [hep-ex].
  7. G. Kasieczka, T. Plehn, M. Russell and T. Schell, “Deep-learning Top Taggers or The End of QCD?,” JHEP 1705, 006 (2017) doi:10.1007/JHEP05(2017)006 [arXiv:1701.08784 [hep-ph]].
  8. L. de Oliveira, M. Paganini and B. Nachman, “Learning Particle Physics by Example: Location-Aware Generative Adversarial Networks for Physics Synthesis,” arXiv:1701.05927 [stat.ML].