Source author record

Roberto Vega

Roberto Vega appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

hep-ph hep-ex Machine Learning Computer Vision physics.acc-ph

Catalog footprint

What is connected

6works

5topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2023arXiv

The International Linear Collider: Report to Snowmass 2021

The International Linear Collider (ILC) is on the table now as a new global energy-frontier accelerator laboratory taking data in the 2030s. The ILC addresses key questions for our current understanding of particle physics. It is based on a proven accelerator technology. Its experiments will challenge the Standard Model of particle physics and will provide a new window to look beyond it. This document brings the story of the ILC up to date, emphasizing its strong physics motivation, its readiness for construction, and the opportunity it presents to the US and the global particle physics community.

preprint2022arXiv

Domain-shift adaptation via linear transformations

A predictor, $f_A : X \to Y$, learned with data from a source domain (A) might not be accurate on a target domain (B) when their distributions are different. Domain adaptation aims to reduce the negative effects of this distribution mismatch. Here, we analyze the case where $P_A(Y\ |\ X) \neq P_B(Y\ |\ X)$, $P_A(X) \neq P_B(X)$ but $P_A(Y) = P_B(Y)$; where there are affine transformations of $X$ that makes all distributions equivalent. We propose an approach to project the source and target domains into a lower-dimensional, common space, by (1) projecting the domains into the eigenvectors of the empirical covariance matrices of each domain, then (2) finding an orthogonal matrix that minimizes the maximum mean discrepancy between the projections of both domains. For arbitrary affine transformations, there is an inherent unidentifiability problem when performing unsupervised domain adaptation that can be alleviated in the semi-supervised case. We show the effectiveness of our approach in simulated data and in binary digit classification tasks, obtaining improvements up to 48% accuracy when correcting for the domain shift in the data.

preprint2021arXiv

Sample Efficient Learning of Image-Based Diagnostic Classifiers Using Probabilistic Labels

Deep learning approaches often require huge datasets to achieve good generalization. This complicates its use in tasks like image-based medical diagnosis, where the small training datasets are usually insufficient to learn appropriate data representations. For such sensitive tasks it is also important to provide the confidence in the predictions. Here, we propose a way to learn and use probabilistic labels to train accurate and calibrated deep networks from relatively small datasets. We observe gains of up to 22% in the accuracy of models trained with these labels, as compared with traditional approaches, in three classification tasks: diagnosis of hip dysplasia, fatty liver, and glaucoma. The outputs of models trained with probabilistic labels are calibrated, allowing the interpretation of its predictions as proper probabilities. We anticipate this approach will apply to other tasks where few training instances are available and expert knowledge can be encoded as probabilities.

preprint2018arXiv

The Supersymmetric Georgi-Machacek Model

We show that the well known Georgi-Machacek (GM) model can be realized as a limit of the recently constructed Supersymmetric Custodial Higgs Triplet Model (SCTM) which in general contains a significantly more complex scalar spectrum. We dub this limit of the SCTM, which gives a weakly coupled origin for the GM model at the electroweak scale, the Supersymmetric GM (SGM) model. We derive a mapping between the SGM and GM models using it to show how a supersymmetric origin implies constraints on the Higgs potential in conventional GM model constructions which would generically not be present. We then perform a simplified phenomenological study of diphoton and ZZ signals for a pair of benchmark scenarios to illustrate under what circumstances the GM model can mimic the SGM model and when they should be easily distinguishable.

preprint2014arXiv

Supersymmetric Custodial Higgs Triplets and the Breaking of Universality

Higgs triplet models are known to have difficulties obtaining agreement with electroweak precision data and in particular constraints on the $ρ$ parameter. Either a global $SU(2)_L \otimes SU(2)_R$ symmetry has to be imposed on the scalar potential at the electroweak scale, as done in the well-known Georgi-Machacek (GM) model, or the triplet vacuum expectation values must be very small. We construct a supersymmetric model that can satisfy constraints on the $ρ$ parameter, even if these two conditions are not fulfilled. We supersymmetrize the GM model by imposing the $SU(2)_L \otimes SU(2)_R$ symmetry at a scale $\mathcal M$, which we argue should be at or above the messenger scale, where supersymmetry breaking is transmitted to the observable sector. We show that scales $\mathcal M$ well above 100 TeV and sizable contributions from the triplets to electroweak symmetry breaking can be comfortably accommodated. We discuss the main phenomenological properties of the model and demonstrate that the departure from custodial symmetry at the electroweak scale, due to radiative breaking, can show up at the LHC as a deviation in the `universal' relation for the Higgs couplings to $WW$ and $ZZ$. As a by-product of supersymmetry, we also show that one can easily obtain both large tree-level \emph{and} one loop corrections to the Higgs mass. This allows for stops that can be significantly lighter and with smaller mixing than those needed in the MSSM.

preprint1993arXiv

Constraints on CP violation in the Higgs sector from the $ρ$ parameter

We discuss the relation between the CP symmetry and the custodial $SU(2)$ symmetry in the Higgs sector. In particular, we show that CP violation in the Higgs-gauge sector is allowed only if the custodial $SU(2)$ symmetry is broken. We exploit these facts to constrain CP violation using the experimental bounds on $ρ-1$. CP nonconservation in the Higgs-fermion interactions can also be constrained in a similar way although a possible exception is pointed out.