Conveners
Mini-workshop on Machine Learning for Particle Physics
- Tommaso Dorigo (Universita e INFN, Padova (IT))
Mini-workshop on Machine Learning for Particle Physics
- There are no conveners in this block
Mini-workshop on Machine Learning for Particle Physics
- There are no conveners in this block
Mini-workshop on Machine Learning for Particle Physics
- There are no conveners in this block
One of the main limitations in particle physics analyses in which the signal selection is based on machine learning is the understanding of the implications of systematic uncertainties. The usual approach consisting in the training with samples ignoring systematic effects and estimating their contribution to the magnitudes measured on modified test samples. We propose here an alternative...
Many searches at the LHC experiments target topologies with three or more invisible particles in the final state. The reconstruction of the full event kinematics is in general not possible even using the information provided by the missing transverse momentum or by the constraints based on the presence of known-mass resonances in the decay chain process. On the other hand, the space of...
In heavy-ion collisions at large particle colliders, such as LHC or RHIC, heavy-flavour (charm and beauty) quarks are produced mainly through initial hard scatterings. Therefore, they can serve as probes of the properties of the hot medium created in such collisions. Hadrons, that contain such quarks, could not be directly detected, thus they are measured via reconstruction of their decay...
A challenge for future particle-physics experiments at the high-energy frontier is the precise measurement of muon momenta at very high energy. In this work we discuss the feasibility of an entirely new avenue for the measurement of the energy of muons based on their radiative losses in a dense, finely segmented calorimeter. We demonstrate with an idealised calorimeter layout, how spatial and...
Searches for pairs of Higgs bosons will be, in all likelihood, the best tools to precisely measure the Higgs boson self-coupling $\lambda_{hhh}$ in future colliders. We study various strategies for the $hh\to b \bar{b} b \bar{b}$ search in the HL-LHC era with focus on constraining $\lambda_{hhh}$. We implement a machine-learning-based approach to separate signal and background and apply...
We study how to use Deep Variational Autoencoders for a fast simulation of jets of particles at the LHC. We represent jets as a list of constituents, characterized by their momenta. Starting from a simulation of the jet before detector effects, we train a Deep Variational Autoencoder to return the corresponding list of constituents after detection. Doing so, we bypass both the detector...
PANDA is a hadron physics research detector at the FAIR facility in Darmstadt, using antiproton beams with momenta between 1.5 and 15 GeV/c interacting with fixed proton targets. From the scientific requirements, the high-performance of electromagnetic calorimeters (EMC) is of utmost importance for the success of the PANDA experiment. Excellent identification and reconstruction of...
Deep machine learning methods have been studied for the PANDA software trigger with data sets from full Monte Carlo simulation using PandaRoot. Following the first comparison of multiclass and binary classification, the binary classification has been selected because of higher signal efficiencies. In total seven neural network types have been compared and the residual convolutional neural...
The Phase-II upgrade of the LHC will increase its instantaneous luminosity by a factor of 7 leading to the High Luminosity LHC (HL-LHC). At the HL-LHC, the number of proton-proton collisions in one bunch crossing (called pileup) increases significantly, putting more stringent requirements on the LHC detectors electronics and real-time data processing capabilities.
The ATLAS Liquid Argon...
In modern neural networks, supervised learning is implemented as minimization of a loss function that typically represents an estimate of the prediction error on the training samples.The gradient of the loss function is traversed in steps towards the minimum, and at each step the prediction error is propagated backwards to all the network weights.The gradient steps are computed using the loss...
In the collider physics searches, missing values can occur if some of the final state particles are not present in all the events. The electroweak production of the $Z\gamma jj$ – a good probe for the electroweak symmetry breaking – is an example of a process with such final state. Third jet parameters are known to be good at distinguishing it from its’ main background – QCD $Z\gamma jj$...
Searches for new physics at the LHC typically focus on well-specified new physics models. However, this may leave interesting potential signals untested. In this presentation, we describe a search method that does not assume a specific form for the searched distributions. The method is based on a scan of the copula space of multidimensional features of collider events. The performances are...
