Punzi-loss Abudinén, F.; Bertemes, M.; Bilokin, S. ...
The European physical journal. C, Particles and fields,
2022/2, Letnik:
82, Številka:
2
Journal Article
Recenzirano
Odprti dostop
We present the novel implementation of a non-differentiable metric approximation and a corresponding loss-scheduling aimed at the search for new particles of unknown mass in high energy physics ...experiments. We call the loss-scheduling, based on the minimisation of a figure-of-merit related function typical of particle physics, a Punzi-loss function, and the neural network that utilises this loss function a Punzi-net. We show that the Punzi-net outperforms standard multivariate analysis techniques and generalises well to mass hypotheses for which it was not trained. This is achieved by training a single classifier that provides a coherent and optimal classification of all signal hypotheses over the whole search space. Our result constitutes a complementary approach to fully differentiable analyses in particle physics. We implemented this work using PyTorch and provide users full access to a public repository containing all the codes and a training example.
Punzi-loss Abudinén, F; Bertemes, M; Bilokin, S ...
European physical journal. C, Particles and fields,
02/2022, Letnik:
82, Številka:
2
Journal Article
Recenzirano
Odprti dostop
We present the novel implementation of a non-differentiable metric approximation and a corresponding loss-scheduling aimed at the search for new particles of unknown mass in high energy physics ...experiments. We call the loss-scheduling, based on the minimisation of a figure-of-merit related function typical of particle physics, a Punzi-loss function, and the neural network that utilises this loss function a Punzi-net. We show that the Punzi-net outperforms standard multivariate analysis techniques and generalises well to mass hypotheses for which it was not trained. This is achieved by training a single classifier that provides a coherent and optimal classification of all signal hypotheses over the whole search space. Our result constitutes a complementary approach to fully differentiable analyses in particle physics. We implemented this work using PyTorch and provide users full access to a public repository containing all the codes and a training example.
We report the first search for a non-standard-model resonance decaying into $\tau$ pairs in $e^{+}e^{-}\rightarrow \mu^{+}\mu^{-} \tau^+\tau^-$ events in the 3.6-10 GeV/$c^{2}$ mass range. We use a ...62.8 fb$^{-1}$ sample of $e^+e^-$ collisions collected at a center-of-mass energy of 10.58 GeV by the Belle II experiment at the SuperKEKB collider. The analysis probes three different models predicting a spin-1 particle coupling only to the heavier lepton families, a Higgs-like spin-0 particle that couples preferentially to charged leptons (leptophilic scalar), and an axion-like particle, respectively. We observe no evidence for a signal and set exclusion limits at 90% confidence level on the product of cross section and branching fraction into $\tau$ pairs, ranging from 0.7 fb to 24 fb, and on the couplings of these processes. We obtain world-leading constraints on the couplings for the leptophilic scalar model for masses above 6.5 GeV/$c^2$ and for the axion-like particle model over the entire mass range.
The $L_{\mu}-L_{\tau}$ extension of the standard model predicts the existence of a lepton-flavor-universality-violating $Z^{\prime}$ boson that couples only to the heavier lepton families. We search ...for such a $Z^\prime$ through its invisible decay in the process $e^+ e^- \to \mu^+ \mu^- Z^{\prime}$. We use a sample of electron-positron collisions at a center-of-mass energy of 10.58GeV collected by the Belle II experiment in 2019-2020, corresponding to an integrated luminosity of 79.7fb$^{-1}$. We find no excess over the expected standard-model background. We set 90$\%$-confidence-level upper limits on the cross section for this process as well as on the coupling of the model, which ranges from $3 \times 10^{-3}$ at low $Z^{\prime}$ masses to 1 at $Z^{\prime}$ masses of 8$GeV/c^{2}$.
We study the processes $e^+e^-\to\omega\chi_{bJ}(1P)$ ($J$ = 0, 1, or 2) using samples at center-of-mass energies $\sqrt{s}$ = 10.701, 10.745, and 10.805 GeV, corresponding to 1.6, 9.8, and 4.7 ...fb$^{-1}$ of integrated luminosity, respectively. These data were collected with the Belle II detector during special operations of the SuperKEKB collider above the $\Upsilon(4S)$ resonance. We report the first observation of $\omega\chi_{bJ}(1P)$ signals at $\sqrt{s}$ = 10.745 GeV. By combining Belle II data with Belle results at $\sqrt{s}$ = 10.867 GeV, we find energy dependencies of the Born cross sections for $e^+e^-\to \omega\chi_{b1,b2}(1P)$ to be consistent with the shape of the $\Upsilon(10753)$ state. These data indicate that the internal structures of the $\Upsilon(10753)$ and $\Upsilon(10860)$ states may differ. Including data at $\sqrt{s}$ = 10.653 GeV, we also search for the bottomonium equivalent of the $X(3872)$ state decaying into $\omega\Upsilon(1S)$. No significant signal is observed for masses between 10.45 and 10.65 GeV/$c^2$.
We describe the principles and performance of the first-level ("L1") hardware track trigger of Belle II, based on neural networks. The networks use as input the results from the standard Belle II ...trigger, which provides "2D" track candidates in the plane transverse to the electron-positron beams. The networks then provide estimates for the origin of the 2D track candidates in direction of the colliding beams ("\(z\)-vertex"), as well as their polar emission angles \(\theta\). Given the \(z\)-vertices of the "neural" tracks allows identifying events coming from the collision region (\(z \approx 0\)), and suppressing the overwhelming background from outside by a suitable cut \(d\). Requiring \(|z| < d\) for at least one neural track in an event with two or more 2D candidates will set an L1 trigger. The networks also enable a minimum bias trigger, requiring a single 2D track candidate validated by a neural track with a momentum larger than 0.7 GeV in addition to the \(|z|\) condition. The momentum of the neural track is derived with the help of the polar angle \(\theta\).
We present the novel implementation of a non-differentiable metric approximation and a corresponding loss-scheduling aimed at the search for new particles of unknown mass in high energy physics ...experiments. We call the loss-scheduling, based on the minimisation of a figure-of-merit related function typical of particle physics, a Punzi-loss function, and the neural network that utilises this loss function a Punzi-net. We show that the Punzi-net outperforms standard multivariate analysis techniques and generalises well to mass hypotheses for which it was not trained. This is achieved by training a single classifier that provides a coherent and optimal classification of all signal hypotheses over the whole search space. Our result constitutes a complementary approach to fully differentiable analyses in particle physics. We implemented this work using PyTorch and provide users full access to a public repository containing all the codes and a training example.
Belle II Executive Summary Asner, D M; Atmacan, H; Banerjee, Sw ...
arXiv (Cornell University),
03/2022
Paper, Journal Article
Odprti dostop
Belle II is a Super \(B\) Factory experiment, expected to record 50 ab\(^{-1}\) of \(e^+e^-\) collisions at the SuperKEKB accelerator over the next decade. The large samples of \(B\) mesons, charm ...hadrons, and tau leptons produced in the clean experimental environment of \(e^+e^-\) collisions will provide the basis of a broad and unique flavor-physics program. Belle II will pursue physics beyond the Standard Model in many ways, for example: improving the precision of weak interaction parameters, particularly Cabibbo-Kobayashi-Maskawa (CKM) matrix elements and phases, and thus more rigorously test the CKM paradigm, measuring lepton-flavor-violating parameters, and performing unique searches for missing-mass dark matter events. Many key measurements will be made with world-leading precision.
Hemiparesis resulting from a stroke has a direct impact on patients' daily activities. New approaches for motor rehabilitation include Serious Games (SG) because they include (in a motivating way) ...the three fundamental elements for rehabilitation: intensive, repetitive and task-oriented training. This study aims to evaluate the therapeutic effects of a biomedical SG and a scoring system developed for lower limb motor rehabilitation of hemiparetic stroke patients. The SG was inspired by the classic videogame called Pong, where the goal is to control a tennis racquet, but using muscular strength. A knee extensor apparatus was adapted with a load cell and mechanical adjustments for measuring the muscular strength of the quadriceps femoris (QFG) and hamstrings (HSG). A scoring system was proposed to evaluate muscular control. Eleven hemiparetic stroke patients participated in an exercise program using the SG twice a week for ten weeks and only the paretic side was trained. Significant Effect Sizes (d) were found for QFG strength (d = 0.5; p = 0.021), QFG control (d = 1.1; p <; 0.001), HSG strength (d = 1.1; p = 0.001), HSG control (d = 1.5; p = 0.003), functional mobility (d = 0.3; p <; 0.001), gait speed (d = 0.4; p = 0.007) and motor recovery (d = 1.0; p <; 0.001). Results indicate that the intervention of a SG with both proper apparatus and evaluation system may effectively promote lower limb motor rehabilitation of hemiparetic stroke patients.
ABSTRACT
We present results from the first public data release of the MaNGA-ARO Survey of CO Targets (MASCOT), focusing our study on galaxies whose star formation rates and stellar masses place them ...below the ridge of the star-forming main sequence. In optically selected type 2 AGN/low-ionization nuclear emission regions (LINERs)/Composites, we find an empirical relation between gas-phase metallicity gradients ∇Z and global molecular gas depletion times $t_{\rm dep} = M_{\rm H_2} /{\rm SFR}$ with ‘more quenched’ systems showing flatter/positive gradients. Our results are based on the O3N2 metallicity diagnostic (applied to star-forming regions within a given galaxy), which was recently suggested to also be robust against emission by diffuse ionized gas (DIG) and LINERs. We conduct a systematic investigation into possible drivers of the observed ∇Z − tdep relation (ouflows, gas accretion, in situ star formation, mergers, and morphology). We find a strong relation between ∇Z or tdep and centralized outflow strength traced by the O iii velocity broadening. We also find signatures of suppressed star formation in the outskirts in AGN-like galaxies with long depletion times and an enhancement of metals in the outer regions. We find no evidence of inflows impacting the metallicity gradients, and none of our results are found to be significantly affected by merger activity or morphology. We thus conclude that the observed ∇Z–tdep relation may stem from a combination of metal redistribution via weak feedback, and a connection to in situ star formation via a resolved mass-metallicity–SFR relation.