486 Hits in 1.3 sec

Similarity Hashing and Learning for Tracks Reconstruction

Sabrina Amrouche, Tobias Golling, Moritz Kiehn, Andreas Salzburger
2019 Zenodo  
At the High Luminosity Large Hadron Collider (HL-LHC), many proton-proton collisions happen during a single bunch crossing. This leads on average to tens of thousands of particles emerging from the interaction region. Two major factors impede finding charged particle trajectories from measured hits in the tracking detectors. First, deciding whether a given set of hits was produced by a common particle is an under-specified task. State-of-the-art reconstruction models usually tackle this issue
more » ... a so-called track following only at a later stage after considering many hits. Second, assuming a nearly perfect hit-particle decision function, constructing possible hit combinations to their compatibility using this decision function is a combinatorial problem. Thus, the traditional approach will grow exponentially as the number of simultaneous collisions increase at the HL-LHC and pose a major computational challenge. We propose a framework for Similarity Hashing and Learning for Track Reconstruction (SHLTR) where multiple small regions of the detector are reconstructed in parallel with minimal fake rate. We use hashing techniques to separate the detector search space into buckets. The particle purity of these buckets, i.e. how many hits from the same particle are contained, is increased using locality sensitivity in feature space where per-hit features beyond just its position are considered. The bucket size is sufficiently small to significantly reduce the complexity of track reconstruction within the buckets or regions. A neural network selects valid combinations in the buckets and builds up full trajectories by connected components search independently of global positions of the hits and detector geometry. The whole process occurs simultaneously in the multiple regions of the detector and curved particles are found by allowing buckets to overlap. We present first results of such a track reconstruction chain including efficiency, fake estimates, and computational performances in µ=200 datasets.
doi:10.5281/zenodo.3599393 fatcat:bjezwy6edfdctosvk7lpwzzhbi

Search for New Physics Involving Top Quarks at ATLAS [article]

Tobias Golling, for the ATLAS Collaboration
2011 arXiv   pre-print
Two searches for new phenomena involving top quarks are presented: a search for a top partner in ttbar events with large missing transverse momentum, and a search for ttbar resonances in proton-proton collisions at a center-of-mass energy of 7 TeV. The measurements are based on 35 pb^-1 and 200 pb^-1 of data collected with the ATLAS detector at the LHC in 2010 and 2011, respectively. No evidence for a signal is observed. The first limits from the LHC are established on the mass of a top
more » ... excluding a mass of 275 GeV for a neutral particle mass less than 50 GeV and a mass of 300 GeV for a neutral particle mass less than 10 GeV. Using the reconstructed ttbar mass spectrum, limits are set on the production cross-section times branching ratio to ttbar for narrow and wide resonances. For narrow Z' models, the observed 95% C.L. limits range from approximately 38 pb to 3.2 pb for masses going from m_Z' = 500 GeV to m_Z' = 1300 GeV. In Randall-Sundrum models, Kaluza-Klein gluons with masses below 650 GeV are excluded at 95% C.L.
arXiv:1109.6734v1 fatcat:l7x3msuphnc7jnkgiecpyoc3be

Funnels: Exact maximum likelihood with dimensionality reduction [article]

Samuel Klein, John A. Raine, Sebastian Pina-Otey, Slava Voloshynovskiy, Tobias Golling
2021 arXiv   pre-print
Normalizing flows are diffeomorphic, typically dimension-preserving, models trained using the likelihood of the model. We use the SurVAE framework to construct dimension reducing surjective flows via a new layer, known as the funnel. We demonstrate its efficacy on a variety of datasets, and show it improves upon or matches the performance of existing flows while having a reduced latent space size. The funnel layer can be constructed from a wide range of transformations including restricted convolution and feed forward layers.
arXiv:2112.08069v1 fatcat:s5iefbqsk5a4ridwc3qllkvlh4

Information-theoretic stochastic contrastive conditional GAN: InfoSCC-GAN [article]

Vitaliy Kinakh, Mariia Drozdova, Guillaume Quétant, Tobias Golling, Slava Voloshynovskiy
2021 arXiv   pre-print
Conditional generation is a subclass of generative problems where the output of the generation is conditioned by the attribute information. In this paper, we present a stochastic contrastive conditional generative adversarial network (InfoSCC-GAN) with an explorable latent space. The InfoSCC-GAN architecture is based on an unsupervised contrastive encoder built on the InfoNCE paradigm, an attribute classifier and an EigenGAN generator. We propose a novel training method, based on generator
more » ... arization using external or internal attributes every n-th iteration, using a pre-trained contrastive encoder and a pre-trained classifier. The proposed InfoSCC-GAN is derived based on an information-theoretic formulation of mutual information maximization between input data and latent space representation as well as latent space and generated data. Thus, we demonstrate a link between the training objective functions and the above information-theoretic formulation. The experimental results show that InfoSCC-GAN outperforms the "vanilla" EigenGAN in the image generation on AFHQ and CelebA datasets. In addition, we investigate the impact of discriminator architectures and loss functions by performing ablation studies. Finally, we demonstrate that thanks to the EigenGAN generator, the proposed framework enjoys a stochastic generation in contrast to vanilla deterministic GANs yet with the independent training of encoder, classifier, and generator in contrast to existing frameworks. Code, experimental results, and demos are available online at
arXiv:2112.09653v1 fatcat:hm43huaklfhg3obabmxzxatzkm

CURTAINs for your Sliding Window: Constructing Unobserved Regions by Transforming Adjacent Intervals [article]

John Andrew Raine, Samuel Klein, Debajyoti Sengupta, Tobias Golling
2022 arXiv   pre-print
We propose a new model independent technique for constructing background data templates for use in searches for new physics processes at the LHC. This method, called CURTAINs, uses invertible neural networks to parametrise the distribution of side band data as a function of the resonant observable. The network learns a transformation to map any data point from its value of the resonant observable to another chosen value. Using CURTAINs, a template for the background data in the signal window is
more » ... constructed by mapping the data from the side-bands into the signal region. We perform anomaly detection using the CURTAINs background template to enhance the sensitivity to new physics in a bump hunt. We demonstrate its performance in a sliding window search across a wide range of mass values. Using the LHC Olympics dataset, we demonstrate that CURTAINs outperforms other leading approaches which aim to improve the sensitivity of bump hunts. It can be trained on a much smaller range of the invariant mass, and is fully data driven.
arXiv:2203.09470v1 fatcat:i3ycei23und2zkn4hpyu6ep7w4

Generation of data on discontinuous manifolds via continuous stochastic non-invertible networks [article]

Mariia Drozdova, Vitaliy Kinakh, Guillaume Quétant, Tobias Golling, Slava Voloshynovskiy
2021 arXiv   pre-print
The generation of discontinuous distributions is a difficult task for most known frameworks such as generative autoencoders and generative adversarial networks. Generative non-invertible models are unable to accurately generate such distributions, require long training and often are subject to mode collapse. Variational autoencoders (VAEs), which are based on the idea of keeping the latent space to be Gaussian for the sake of a simple sampling, allow an accurate reconstruction, while they
more » ... ence significant limitations at generation task. In this work, instead of trying to keep the latent space to be Gaussian, we use a pre-trained contrastive encoder to obtain a clustered latent space. Then, for each cluster, representing a unimodal submanifold, we train a dedicated low complexity network to generate this submanifold from the Gaussian distribution. The proposed framework is based on the information-theoretic formulation of mutual information maximization between the input data and latent space representation. We derive a link between the cost functions and the information-theoretic formulation. We apply our approach to synthetic 2D distributions to demonstrate both reconstruction and generation of discontinuous distributions using continuous stochastic networks.
arXiv:2112.09646v1 fatcat:n472gnyvgnar3bnmdaazzusfge

Generative modeling for shower simulation in ATLAS

Dalila Salamani, Tobias Golling, Graeme A Stewart, Stefan Gadatsch, Johnny Raine, Aishik Ghosh, David Rousseau, Gilles Louppe, Kyle Stuart Cranmer
2019 Zenodo  
Modeling the physics of a detector's response to particle collisions is one of the most CPU intensive and time consuming aspects of LHC computing. With the upcoming high-luminosity upgrade and the need to have even larger simulated datasets to support physics analysis, the development of new faster simulation techniques but with sufficiently accurate physics performance is required. The current ATLAS fast calorimeter simulation technique, based on parametrizations of the calorimeter response in
more » ... the longitudinal and transverse direction given a single particle's eta and energy, is being updated to higher accuracy including machine learning approaches. Here we report on a prototype using cutting edge machine learning approaches to learn the appropriate detector output response, which are expected to lead to an improved modelling of correlations within showers. The model is trained on a set of fully simulated events and the goal is to generate new outputs. We are studying Variational Auto-Encoders (VAEs) and Generative Adversarial Networks (GANs) to model particle showers in the ATLAS calorimeter. In this study we present an exploratory analysis of both models trained with different optimization tricks and criteria. Thus, our goal is get a deeper understanding of the learning process and how it leads to better improvement of the generation performance.
doi:10.5281/zenodo.3599108 fatcat:jpkxrt4dbnc4pfps5s6ilun2ae

Enhanced Higgs boson coupling to charm pairs

Cédric Delaunay, Tobias Golling, Gilad Perez, Yotam Soreq
2014 Physical Review D  
We show that current Higgs data permit a significantly enhanced Higgs coupling to charm pairs, comparable to the Higgs to bottom pairs coupling in the Standard Model, without resorting to additional new physics sources in Higgs production. With a mild level of the latter current data even allow for the Higgs to charm pairs to be the dominant decay channel. An immediate consequence of such a large charm coupling is a significant reduction of the Higgs signal strengths into the known final states
more » ... as in particular into bottom pairs. This might reduce the visible vector-boson associated Higgs production rate to a level that could compromise the prospects of ever observing it. We however demonstrate that a significant fraction of this reduced signal can be recovered by jet-flavor-tagging targeted towards charm-flavored jets. Finally we argue that an enhanced Higgs to charm pairs coupling can be obtained in various new physics scenarios in the presence of only a mild accidental cancellation between various contributions.
doi:10.1103/physrevd.89.033014 fatcat:mn4jecvtgzhmrjb7qqsec3gbpm

SUPA: A Lightweight Diagnostic Simulator for Machine Learning in Particle Physics [article]

Atul Kumar Sinha, Daniele Paliotta, Bálint Máté, Sebastian Pina-Otey, John A. Raine, Tobias Golling, François Fleuret
2022 arXiv   pre-print
Deep learning methods have gained popularity in high energy physics for fast modeling of particle showers in detectors. Detailed simulation frameworks such as the gold standard Geant4 are computationally intensive, and current deep generative architectures work on discretized, lower resolution versions of the detailed simulation. The development of models that work at higher spatial resolutions is currently hindered by the complexity of the full simulation data, and by the lack of simpler, more
more » ... interpretable benchmarks. Our contribution is SUPA, the SUrrogate PArticle propagation simulator, an algorithm and software package for generating data by simulating simplified particle propagation, scattering and shower development in matter. The generation is extremely fast and easy to use compared to Geant4, but still exhibits the key characteristics and challenges of the detailed simulation. We support this claim experimentally by showing that performance of generative models on data from our simulator reflects the performance on a dataset generated with Geant4. The proposed simulator generates thousands of particle showers per second on a desktop machine, a speed up of up to 6 orders of magnitudes over Geant4, and stores detailed geometric information about the shower propagation. SUPA provides much greater flexibility for setting initial conditions and defining multiple benchmarks for the development of models. Moreover, interpreting particle showers as point clouds creates a connection to geometric machine learning and provides challenging and fundamentally new datasets for the field. The code for SUPA is available at
arXiv:2202.05012v1 fatcat:24ox556mdzhxnnhbnidjvjypam

Variational Autoencoders for Anomalous Jet Tagging [article]

Taoli Cheng, Jean-François Arguin, Julien Leissner-Martin, Jacinthe Pilette, Tobias Golling
2021 arXiv   pre-print
We present a detailed study on Variational Autoencoders (VAEs) for anomalous jet tagging at the Large Hadron Collider. By taking in low-level jet constituents' information, and training with background QCD jets in an unsupervised manner, the VAE is able to encode important information for reconstructing jets, while learning an expressive posterior distribution in the latent space. When using the VAE as an anomaly detector, we present different approaches to detect anomalies: directly comparing
more » ... n the input space or, instead, working in the latent space. In order to facilitate general search approaches such as bump-hunt, mass-decorrelated VAEs based on distance correlation regularization are also studied. We find that the naive mass-decorrelated VAEs fail at maintaining proper detection performance, by assigning higher probabilities to some anomalous samples. To build a performant mass-decorrelated anomalous jet tagger, we propose the Outlier Exposed VAE (OE-VAE), for which some outlier samples are introduced in the training process to guide the learned information. OE-VAEs are employed to achieve two goals at the same time: increasing sensitivity of outlier detection and decorrelating jet mass from the anomaly score. We succeed in reaching excellent results from both aspects. Code implementation of this work can be found at \href{}{Github}.
arXiv:2007.01850v3 fatcat:jxz355tcknesteqqh5adoxuwke

Differential contractile response of critically ill patients to neuromuscular electrical stimulation

Julius J. Grunow, Moritz Goll, Niklas M. Carbon, Max E. Liebl, Steffen Weber-Carstens, Tobias Wollersheim
2019 Critical Care  
Neuromuscular electrical stimulation (NMES) has been investigated as a preventative measure for intensive care unit-acquired weakness. Trial results remain contradictory and therefore inconclusive. As it has been shown that NMES does not necessarily lead to a contractile response, our aim was to characterise the response of critically ill patients to NMES and investigate potential outcome benefits of an adequate contractile response.
doi:10.1186/s13054-019-2540-4 pmid:31506074 pmcid:PMC6737711 fatcat:pgzyaqnm3nd27jitgg7dpek2fe

Turbo-Sim: a generalised generative model with a physical latent space [article]

Guillaume Quétant, Mariia Drozdova, Vitaliy Kinakh, Tobias Golling, Slava Voloshynovskiy
2021 arXiv   pre-print
We present Turbo-Sim, a generalised autoencoder framework derived from principles of information theory that can be used as a generative model. By maximising the mutual information between the input and the output of both the encoder and the decoder, we are able to rediscover the loss terms usually found in adversarial autoencoders and generative adversarial networks, as well as various more sophisticated related models. Our generalised framework makes these models mathematically interpretable
more » ... nd allows for a diversity of new ones by setting the weight of each loss term separately. The framework is also independent of the intrinsic architecture of the encoder and the decoder thus leaving a wide choice for the building blocks of the whole network. We apply Turbo-Sim to a collider physics generation problem: the transformation of the properties of several particles from a theory space, right after the collision, to an observation space, right after the detection in an experiment.
arXiv:2112.10629v2 fatcat:3xzhoybdr5f33kh6upwfhkgrji

Commissioning of the ATLAS pixel detector

Tobias Golling
2009 Nuclear Instruments and Methods in Physics Research Section A : Accelerators, Spectrometers, Detectors and Associated Equipment  
doi:10.1016/j.nima.2009.01.108 fatcat:nndqmilc4reu3jzl2hyqnj55ke

LHC searches for physics beyond the Standard Model with top quarks

Tobias Golling
2013 Journal of Physics, Conference Series  
Searches are presented for physics beyond the Standard Model involving top-quark and related signatures. The results are based on proton-proton collision data corresponding to integrated luminosities between 1 fb-1 and 5 fb-1 collected at a center-of-mass energy of 7 TeV with the ATLAS and CMS detectors at the Large Hadron Collider in 2011. The data are found to be consistent with the Standard Model. The non-observation of a signal is converted to limits at the 95% confidence level on the
more » ... tion cross section times branching ratio and on the masses of the hypothesized new particles for appropriate benchmark models.
doi:10.1088/1742-6596/452/1/012010 fatcat:hiaouor5hjbqtmd6exuguwl2n4

The Tracking Machine Learning Challenge

David Rousseau, Jean-Roch Vlimant, Vincenzo Innocente, Andreas Salzburger, Isabelle Guyon, Sabrina Amrouche, Tobias Golling, Moritz Kiehn, Yetkin Yilmaz, Paolo Calafiura, Steven Farrell, Heather Gray (+7 others)
2019 Zenodo  
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 are provided events with 100k 3D points, and are asked to group the points into tracks; they are also given a 100GB training dataset
more » ... the ground truth. The challenge is run in two phases. The first "Accuracy" phase has run on Kaggle platform from May to August 2018; algorithms were judged judged only on a score related to the fraction of correctly assigned hits. The second "Throughput" phase ran Sep 2018 to March 2019 on Codalab, required code submission; algorithms were then ranked by combining accuracy and speed. The first phase has seen 653 participants, with top performers with innovative approaches (see arXiv:1904.06778). The second phase has recently finished and featured some astonishingly fast solutions. A "grand Finale" workshop will have taken place at CERN early July 2019. The talk will report on the lessons from the TrackML challenge and perspectives
doi:10.5281/zenodo.3599205 fatcat:ta37bipfpzbqhbjo7xfe7jxkdy
« Previous Showing results 1 — 15 out of 486 results