Learning physics at future e−e+ colliders with machine

Experiments at a future e+e− collider will be able to search for new particles with masses below the nominal centre-of-mass energy by analyzing collisions with initial-state radiation (radiative return). We show that machine learning methods that use imperfect or missing training labels can achieve sensitivity to generic new particle production in radiative return events. In addition to presenting an application of the classification without labels (CWoLa) search method in e+e− collisions, our study combines weak supervision with variable-dimensional information by deploying a deep sets neural network architecture. We have also investigated some of the experimental aspects of anomaly detection in radiative return events and discuss these in the context of future detector design.

Section: Jhep04(2022)156mentioning

confidence: 99%

High-dimensional anomaly detection with radiative return in e+e− collisions

Gonski

Lai

Nachman

et al. 2022

“…hadronic modes. As demonstrated in [62], the tool of DNN is powerful in synergizing the kinematic messages at hadron level, and hence may significantly improve many of the baseline precisions presented in the literatures. At last, we have strong motivation to extend this study to the b → cτ ν measurements at the future Z factories.…”

Section: A Detailed Cut Flows and Tera-z Yields For The B → Sτmentioning

confidence: 99%

b → sτ+τ− physics at future Z factories

Liu

2021

Self Cite

b → sτ+τ− measurements are highly motivated for addressing lepton-flavor-universality (LFU)-violating puzzles such as $$ {R}_{K^{\left(\ast \right)}} $$ R K ∗ anomalies. The anomalies of $$ {R}_{D^{\left(\ast \right)}} $$ R D ∗ and RJ/ψ further strengthen their necessity and importance, given that the LFU-violating hints from both involve the third-generation leptons directly. Z factories at the future e−e+ colliders stand at a great position to conduct such measurements because of their relatively high production rates and reconstruction efficiencies for B mesons at the Z pole. To fully explore this potential, we pursue a dedicated sensitivity study in four b → sτ+τ− benchmark channels, namely B0 → K*0τ+τ−, Bs → ϕτ+τ−, B+ → K+τ+τ− and Bs → τ+τ−, at the future Z factories. We develop a fully tracker-based scheme for reconstructing the signal B mesons and introduce a semi-quantitative method for estimating their major backgrounds. The simulations indicate that branching ratios of the first three channels can be measured with a precision ∼ $$ \mathcal{O} $$ O (10−7− 10−6) and that of Bs → τ+τ− with a precision ∼ $$ \mathcal{O} $$ O (10−5) at Tera-Z. The impacts of luminosity and tracker resolution on the expected sensitivities are explored. The interpretations of these results in effective field theory are also presented.

From optimal observables to machine learning: an effective-field-theory analysis of e+e− → W+W− at future lepton colliders

Chai,

Gu,

2024

We apply machine-learning techniques to the effective-field-theory analysis of the e+e− → W+W− processes at future lepton colliders, and demonstrate their advantages in comparison with conventional methods, such as optimal observables. In particular, we show that machine-learning methods are more robust to detector effects and backgrounds, and could in principle produce unbiased results with sufficient Monte Carlo simulation samples that accurately describe experiments. This is crucial for the analyses at future lepton colliders given the outstanding precision of the e+e− → W+W− measurement (~ 10−4 in terms of anomalous triple gauge couplings or even better) that can be reached. Our framework can be generalized to other effective-field-theory analyses, such as the one of e+e− → $$ t\overline{t} $$ t t ¯ or similar processes at muon colliders.