Login

3rd IML Machine Learning Workshop

15–18 Apr 2019
CERN
Europe/Zurich timezone
There is a live webcast for this event.

Contribution List

79 out of 79 displayed
Export to PDF
  1. 14. Welcome
    Steven Randolph Schramm (Universite de Geneve (CH))
    15/04/2019, 09:00
  2. 15. Conceptual overview of ML in HEP
    Dr Sergei Gleyzer (University of Florida (US))
    15/04/2019, 09:20

    This talk is intended to provide a conceptual overview of how ML is used in particle physics, in order to provide a basic level of understanding for all attendees such that people can follow the subsequent talks. Detailed tutorials on the application of ML techniques will instead take place on Friday.

    Go to contribution page
  3. 16. Future areas of focus for ML in particle physics
    Kyle Stuart Cranmer (New York University (US))
    15/04/2019, 10:05
  4. 17. On the Statistical Mechanics and Information Theory of Deep Learning for Particle Physicists
    Prof. Naftali Tishby (Hebrew University of Jerusalem)
    15/04/2019, 11:15

    This is an introduction to Deep Learning, for Particle Physicists. These techniques can provide excellent performance for separating different categories of events, pattern recognition, etc. In order to achieve this it is highly desirable for users to understand the basic concepts underlying their operation. The Statistical Mechanics and Information Theory aspects of this will be addressed in...

    Go to contribution page
  5. 10. Introduction
    Markus Stoye (CERN)
    15/04/2019, 14:00
  6. 9. Amazon AI: Tensor and Higher-Order Generalizations of the GSVD with Applications to Personalized Cancer Medicine
    Dr Priya Ponnapalli (Amazon AI)
    15/04/2019, 14:20

    The number of high-dimensional datasets recording multiple aspects of interrelated phenomena is increasing in many areas, from medicine to finance. This drives the need for mathematical frameworks that can simultaneously identify the similar and dissimilar among multiple matrices and tensors, and thus create a single coherent model from the multiple datasets. The generalized singular value...

    Go to contribution page
  7. 11. Google DeepMind: Compressing neural networks
    Dr Tim Genewein (DeepMind)
    15/04/2019, 15:05

    Empirically it has been observed numerous times that trained neural networks often have high degrees of parameter-redundancy. It remains an open theoretical question why this parameter redundancy cannot be reduced before training by using smaller neural networks. On the other hand, the recent scientific literature reports a plethora of practical methods to "compress" neural networks during or...

    Go to contribution page
  8. 12. B12 Consulting: Big value out of small data
    Dr Michel Herquet (B12)
    15/04/2019, 16:30

    Machine Learning (and especially Deep Learning) algorithms often require large amounts of data to accomplish their tasks. However, a common problem when such approaches are applied in business contexts is that only relatively small datasets are initially accessible, leading to a fundamental question: how to apply ML tools when there is apparently not enough data available? In this talk, I will...

    Go to contribution page
  9. 13. Industry panel
    Dr Michel Herquet (B12), Dr Tim Genewein (DeepMind)
    15/04/2019, 17:05
  10. 18. Daily announcements
    16/04/2019, 09:00
  11. 37. Machine Learning Uncertainties with Adversarial Neural Networks
    Dr Peter Galler (University of Glasgow)
    16/04/2019, 09:05

    Machine learning is a powerful tool to reveal and exploit correlations in a multi-dimensional parameter space. Making predictions from such correlations is a highly non-trivial task, in particular when the details of the underlying dynamics of a theoretical model are not fully understood. Uncertainties in the training data add towards the complexity of performing machine learning tasks such as...

    Go to contribution page
  12. 49. Decoding Physics Information in DNNs
    Taoli Cheng (University of Montreal)
    16/04/2019, 09:25

    A more dedicated study on the information flow in DNNs will help us understand their behaviour and the deep connection between DNN models and the corresponding tasks. Taking into account our well-established physics analysis framework (observable-based), we present a novel way to interpret DNNs results for HEP, which not only gives a clear physics picture but also inspires interfaces with the...

    Go to contribution page
  13. 43. Learning Invariant Representations using Mutual Information Regularization
    Mr Justin Tan (University of Melbourne)
    16/04/2019, 09:45

    Invariance of learned representations of neural networks against certain sensitive attributes of the input data is a desirable trait in many modern-day applications of machine learning, such as precision measurements in experimental high-energy physics and enforcing algorithmic fairness in the social and financial domain. We present a method for enforcing this invariance through regularization...

    Go to contribution page
  14. 51. Neural networks for the abstraction of the physical symmetries in the nature
    Wonsang Cho (Seoul National University)
    16/04/2019, 10:05

    Neural networks are so powerful universal approximator of complicated patterns in large-scale data, leading the explosive developments of AI in terms of deep learning. However, in many cases, usual neural networks are trained to possess poor level of abstraction, so that the model's predictability and generalizability can be quite unstable, depending on the quality and amount of the data used...

    Go to contribution page
  15. 44. Containers for Machine Learning in HEP
    Matthew Feickert (Southern Methodist University (US))
    16/04/2019, 10:25

    Physicists want to use modern open source machine learning tools developed by industry for machine learning projects and analyses in high energy physics. The software environment that a physicist prototypes, tests, and runs these projects in is ideally the same regardless of compute site (be it their laptop or on the GRID). However, historically it has been difficult to find compute sites that...

    Go to contribution page
  16. 39. The information theory of Deep Learning
    Prof. Naftali Tishby (Hebrew University of Jerusalem)
    16/04/2019, 11:15
  17. 21. Novelty Detection Meets Collider Physics
    Ms Ying-Ying Li (HKUST)
    16/04/2019, 14:00

    Novelty detection is the machine learning task to recognize data, which belong to an unknown pattern. Complementary to supervised learning, it allows to analyze data model-independently. We demonstrate the potential role of novelty detection in collider physics, using autoencoder-based deep neural network. Explicitly, we develop a set of density-based novelty evaluators, which are sensitive to...

    Go to contribution page
  18. 80. Q&A with Prof. Naftali Tishby
    Prof. Naftali Tishby (Hebrew University of Jerusalem)
    16/04/2019, 14:15
  19. 31. Uncertain Networks
    Jennifer Thompson (ITP Heidelberg)
    16/04/2019, 14:30

    Machine learning methods are being increasingly and successfully applied to many different physics problems. However, currently uncertainties in machine learning methods are not modelled well, if at all. In this talk I will discuss how using Bayesian neural networks can give us a handle on uncertainties in machine learning. I will use tagging tops vs. QCD as an example of how these networks...

    Go to contribution page
  20. 28. Exploring SMEFT in VH channel with Machine Learning
    Charanjit Kaur Khosa
    16/04/2019, 15:00

    We use Machine Learning(ML) techniques to exploit kinematic information in VH, the production of a Higgs in association with a massive vector boson. We parametrize the effect of new physics in terms of the SMEFT framework. We find that the use of a shallow neural network allows us to dramatically increase the sensitivity to deviations in VH respect to previous estimates. We also discuss the...

    Go to contribution page
  21. 24. The Tracking Machine Learning challenge
    David Rousseau (LAL-Orsay, FR)
    16/04/2019, 15:30

    The HL-LHC will see ATLAS and CMS see proton bunch collisions reaching track multiplicity up to 10.000 charged tracks per event. Algorithms need to be developed to harness the increased combinatorial complexity. To engage the Computer Science community to contribute new ideas, we have organized a Tracking Machine Learning challenge (TrackML). Participants were provided events with 100k 3D...

    Go to contribution page
  22. 1. Gauge Fields in Deep Learning
    16/04/2019, 16:30

    Invited colloquium by Prof. Dr. Max Welling

    Part of the IML workshop, but listed under the CERN colloquium category.

    Joint work with Taco Cohen, Maurice Weiler and Berkay Kicanaoglu

    ABSTRACT:
    Gauge field theory is the foundation of modern physics, including general relativity and the standard model of physics. It describes how a theory of physics should transform under symmetry...

    Go to contribution page
  23. 19. Daily announcements
    17/04/2019, 09:00
  24. 50. Applying Generative Models to Scientific Research
    Fedor Ratnikov (Yandex School of Data Analysis (RU))
    17/04/2019, 09:05

    Surrogate generative models demonstrate extraordinary progress in current years. Although most applications are dedicated to image generation and similar commercial
    goals, this approach is also very promising for natural sciences, especially for tasks like fast event simulation in HEP experiments. However, application of such generative models to scientific research implies specific...

    Go to contribution page
  25. 47. DijetGAN: A Generative-Adversarial Network Approach for the Simulation of QCD Dijet Events at the LHC
    Serena Palazzo (The University of Edinburgh (GB))
    17/04/2019, 09:35

    In this talk, I will present a Generative-Adversarial Network (GAN) based on convolutional neural networks that is used to simulate the production of pairs of jets at the LHC. The GAN is trained on events generated using MadGraph5 + Pythia8, and Delphes3 fast detector simulation. A number of kinematic distributions both at Monte Carlo truth level and after the detector simulation can be...

    Go to contribution page
  26. 45. High Granularity Calorimeter Simulation using Generative Adversarial Networks
    Gul Rukh Khattak (University of Peshawar (PK))
    17/04/2019, 09:55

    High Energy Physics simulation typically involves Monte Carlo method. Today >50% of WLCG resources are used for simulation that will increase further as detector granularity and luminosity increase. Machine learning has been very successful in the field of image recognition and generation. We have explored image generation techniques for speeding up HEP detector simulation. Calorimeter...

    Go to contribution page
  27. 41. Deep generative models for fast shower simulation in ATLAS
    Aishik Ghosh (Centre National de la Recherche Scientifique (FR))
    17/04/2019, 10:15

    The extensive physics program of the ATLAS experiment at the Large Hadron Collider (LHC) relies on large scale and high fidelity simulation of the detector response to particle interactions. Current full simulation techniques using Geant4 provide accurate modeling of the underlying physics processes, but are inherently resource intensive. In light of the high-luminosity upgrade of the LHC and...

    Go to contribution page
  28. 46. Fast Simulation Using Generative Adversarial Network in LHCB
    Artem Maevskiy (National Research University Higher School of Economics (RU))
    17/04/2019, 11:05

    LHCb is one of the major experiments operating at the Large Hadron Collider at CERN. The richness of the physics program and the increasing precision of the measurements in LHCb lead to the need of ever larger simulated samples. This need will increase further when the upgraded LHCb detector will start collecting data in the LHC Run 3. Given the computing resources pledged for the production...

    Go to contribution page
  29. 22. Model-Assisted GANs for the optimisation of simulation parameters and as an algorithm for fast Monte Carlo production
    Mr Saul Alonso Monsalve (CERN)
    17/04/2019, 11:25

    We propose and demonstrate the use of a Model-Assisted Generative Adversarial Network to produce simulated images that accurately match true images through the variation of underlying model parameters that describe the image generation process. The generator learns the parameter values that give images that best match the true images. The best match parameter values that produce the most...

    Go to contribution page
  30. 38. Event Generation and Statistical Sampling with Deep Generative Models
    Sydney Otten (Radboud Universiteit Nijmegen)
    17/04/2019, 11:45

    We present a study for the generation of events from a physical process with generative deep learning. To simulate physical processes it is not only important to produce physical events, but also to produce the events with the right frequency of occurrence (density). We investigate the feasibility to learn the event generation and the frequency of occurrence with Generative Adversarial...

    Go to contribution page
  31. 29. LUMIN - a deep learning and data science ecosystem for high-energy physics
    Giles Chatham Strong (LIP Laboratorio de Instrumentacao e Fisica Experimental de Part)
    17/04/2019, 12:05

    [LUMIN][1] aims to become a deep-learning and data-analysis ecosystem for High-Energy Physics, and perhaps other scientific domains in the future. Similar to Keras and fastai it is a wrapper framework for a graph computation library (PyTorch), but includes many useful functions to handle domain-specific requirements and problems. It also intends to provide easy access to to state-of-the-art...

    Go to contribution page
  32. 27. A hybrid deep learning approach to vertexing
    Henry Fredrick Schreiner (University of Cincinnati (US))
    17/04/2019, 14:00

    In the transition to Run 3 in 2021, LHCb will undergo a major luminosity upgrade, going from 1.1 to 5.6 expected visible Primary Vertices (PVs) per event, and will adopt a purely software trigger. This has fueled increased interest in alternative highly-parallel and GPU friendly algorithms for tracking and reconstruction. We will present a novel prototype algorithm for vertexing in the LHCb...

    Go to contribution page
  33. 48. Feature ranking based on subtraction methods
    Paul Glaysher (DESY)
    17/04/2019, 14:20

    The input variables of ML methods in physics analysis are often highly correlated and figuring out which ones are the most important ones for the classification turns out to be a non-trivial tasks. We compare the standard method of TMVA to rank variables with a several newly developed methods based on iterative removal for the use case of a search for top pair associated Higgs production...

    Go to contribution page
  34. 30. ML Techniques for heavy flavour identification in CMS
    Emil Sorensen Bols (Vrije Universiteit Brussel (BE))
    17/04/2019, 14:40

    Jet flavour identification is a fundamental component for the physics program of the LHC-based experiments. The presence of multiple flavours to be identified leads to a multiclass classification problem. Moreover, the classification of boosted jets has acquired an increasing importance in the physics program of CMS. In this presentation we will present the performance on both simulated and...

    Go to contribution page
  35. 40. ParticleNet: Jet Tagging via Particle Clouds
    Huilin Qu (Univ. of California Santa Barbara (US))
    17/04/2019, 15:10

    How to represent a jet is at the core of machine learning on jet physics. Inspired by the notion of point cloud, we propose a new approach that considers a jet as an unordered set of its constituent particles, effectively a "particle cloud". Such particle cloud representation of jets is efficient in incorporating raw information of jets and also explicitly respects the permutation symmetry....

    Go to contribution page
  36. 23. Learning representations of irregular particle-detector geometry with distance-weighted graph networks
    Jan Kieseler (CERN)
    17/04/2019, 15:30

    We explore the possibility of using graph networks to deal with irregular-geometry detectors when reconstructing particles. Thanks to their representation-learning capabilities, graph networks can exploit the detector granularity, while dealing with the event sparsity and the irregular detector geometry. In this context, we introduce two distance-weighted graph network architectures, the...

    Go to contribution page
  37. 25. GroomRL: jet grooming through reinforcement learning
    Frederic Alexandre Dreyer (Oxford)
    17/04/2019, 16:30

    We introduce a novel implementation of a reinforcement learning algorithm which is adapted to the problem of jet grooming, a crucial component of jet physics at hadron colliders. We show that the grooming policies trained using a Deep Q-Network model outperform state-of-the-art tools used at the LHC such as Recursive Soft Drop, allowing for improved resolution of the mass of boosted objects....

    Go to contribution page
  38. 42. NeuralRinger: An Ensemble of Neural Networks Fed from Calorimeter Ring Sums for Triggering on Electrons
    Werner Spolidoro Freund (Federal University of of Rio de Janeiro (BR))
    17/04/2019, 16:50

    In 2017, the ATLAS experiment implemented an ensemble of neural networks (NeuralRinger algorithm) dedicated to reduce the latency of the first, fast, online software (HLT) selection stage for electrons with transverse energy above 15 GeV. In order to minimize detector response and shower development fluctuations, and being inspired in the ensemble of likelihood models currently operating in...

    Go to contribution page
  39. 26. Fast Deep Learning on FPGAs for the Phase-II L0 Muon Barrel Trigger of the ATLAS Experiment
    Luigi Sabetta (Sapienza Universita e INFN, Roma I (IT))
    17/04/2019, 17:10

    The Level-0 Muon Trigger system of the ATLAS experiment will undergo a full upgrade for HL-LHC to stand the challenging performances requested with the increasing instantaneous luminosity. The upgraded trigger system foresees to send RPC raw hit data to the off-detector trigger processors, where the trigger algorithms run on new generation of Field-Programmable Gate Arrays (FPGAs). The FPGA...

    Go to contribution page
  40. 20. Close-out
    Steven Randolph Schramm (Universite de Geneve (CH))
    17/04/2019, 17:30
  41. 4. Introduction to the basics of deep learning
    Yannik Alexander Rath (RWTH Aachen University (DE))
    18/04/2019, 09:00
  42. 6. Physics inspired Autonomous Feature Engineering
    Marcel Rieger (RWTH Aachen University (DE))
    18/04/2019, 11:00
  43. 33. Traditional approach to network architecture optimisation
    Andrey Ustyuzhanin (Yandex School of Data Analysis (RU))
    18/04/2019, 12:00
  44. 34. Differentiable architecture search
    Andrey Ustyuzhanin (Yandex School of Data Analysis (RU))
    18/04/2019, 12:30
  45. 35. Bayesian approach to network design
    Andrey Ustyuzhanin (Yandex School of Data Analysis (RU))
    18/04/2019, 15:00
  46. 36. Bayesian dropout explanation and examples
    Andrey Ustyuzhanin (Yandex School of Data Analysis (RU))
    18/04/2019, 15:30
  47. 5. Advanced Generative Adversarial Network Techniques
    Jonas Glombitza (Rheinisch-Westfaelische Tech. Hoch. (DE))
    18/04/2019, 16:45
  48. 2. Coffee break
  49. 3. Coffee break
  50. 52. Talk 1
  51. 55. Talk 1
  52. 60. Talk 1
  53. 66. Talk 1
  54. 53. Talk 2
  55. 56. Talk 2
  56. 61. Talk 2
  57. 67. Talk 2
  58. 54. Talk 3
  59. 57. Talk 3
  60. 62. Talk 3
  61. 68. Talk 3
  62. 58. Talk 4
  63. 69. Talk 4
  64. 74. Talk 4
  65. 75. Talk 4
  66. 59. Talk 5
  67. 63. Talk 5
  68. 70. Talk 5
  69. 76. Talk 5
  70. 64. Talk 6
  71. 71. Talk 6
  72. 65. Talk 7
  73. 72. Talk 7
  74. 73. Talk 8
  75. 77. Talk 8
  76. 78. Talk 9
  77. 32. Towards Fast Displaced Vertex Finding
    Kim Albertsson (Lulea University of Technology (SE))

    Many standard model extensions predict long-lived massive particles that can be detected by looking for displaced decay vertices in the inner detector volume. Current approaches to seek for these events in high-energy particle collisions rely on the presence of additional energetic signatures to make an online selection during data-taking. Enabling trigger-level reconstruction of displaced...

    Go to contribution page
  78. 7. Tutorial, part 1
  79. 8. Tutorial, part 2