Advisor Feed
https://feeds.library.caltech.edu/people/Spiropulu-M/advisor.rss
A Caltech Library Repository Feedhttp://www.rssboard.org/rss-specificationpython-feedgenenWed, 07 Feb 2024 04:47:17 +0000Searches for New Symmetries in pp Collisions with the Razor Kinematic Variables at √s = 7 TeV
https://resolver.caltech.edu/CaltechTHESIS:05072013-174540453
Authors: {'items': [{'email': 'christophersrogan@gmail.com', 'id': 'Rogan-Christopher-Sean', 'name': {'family': 'Rogan', 'given': 'Christopher Sean'}, 'show_email': 'YES'}]}
Year: 2013
DOI: 10.7907/B0MD-F026
<p>The construction and LHC phenomenology of the razor variables M<sub>R</sub>, an event-by-event indicator of the heavy particle mass scale, and R, a dimensionless variable related to the transverse momentum imbalance of events and missing transverse energy, are presented. The variables are used in the analysis of the first proton-proton collisions dataset at CMS (35 pb<sup>-1</sup>) in a search for superpartners of the quarks and gluons, targeting indirect hints of dark matter candidates in the context of supersymmetric theoretical frameworks. The analysis produced the highest sensitivity results for SUSY to date and extended the LHC reach far beyond the previous Tevatron results. A generalized inclusive search is subsequently presented for new heavy particle pairs produced in √s = 7 TeV proton-proton collisions at the LHC using 4.7±0.1 fb<sup>-1</sup> of integrated luminosity from the second LHC run of 2011. The selected events are analyzed in the 2D razor-space of M<sub>R</sub> and R and the analysis is performed in 12 tiers of all-hadronic, single and double leptons final states in the presence and absence of b-quarks, probing the third generation sector using the event heavy-flavor content. The search is sensitive to generic supersymmetry models with minimal assumptions about the superpartner decay chains. No excess is observed in the number or shape of event yields relative to Standard Model predictions. Exclusion limits are derived in the CMSSM framework with gluino masses up to 800 GeV and squark masses up to 1.35 TeV excluded at 95% confidence level, depending on the model parameters. The results are also interpreted for a collection of simplified models, in which gluinos are excluded with masses as large as 1.1 TeV, for small neutralino masses, and the first-two generation squarks, stops and sbottoms are excluded for masses up to about 800, 425 and 400 GeV, respectively. </p>
<p>With the discovery of a new boson by the CMS and ATLAS experiments in the γ-γ and 4 lepton final states, the identity of the putative Higgs candidate must be established through the measurements of its properties. The spin and quantum numbers are of particular importance, and we describe a method for measuring the J<sup>PC</sup> of this particle using the observed signal events in the H to ZZ<sup>*</sup> to 4 lepton channel developed before the discovery. Adaptations of the razor kinematic variables are introduced for the H to WW<sup>*</sup> to 2 lepton/2 neutrino channel, improving the resonance mass resolution and increasing the discovery significance. The prospects for incorporating this channel in an examination of the new boson J<sup>PC</sup> is discussed, with indications that this it could provide complementary information to the H to ZZ<sup>*</sup> to 4 lepton final state, particularly for measuring CP-violation in these decays.</p>https://thesis.library.caltech.edu/id/eprint/7671Multi-Model Inference Ranking and Applications to Physics at the Large Hadron Collider
https://resolver.caltech.edu/CaltechTHESIS:06122014-143656852
Authors: {'items': [{'email': 'valerelambert@gmail.com', 'id': 'Lambert-Valere-Regis-Westbrooke', 'name': {'family': 'Lambert', 'given': 'Valere Regis Westbrooke'}, 'show_email': 'YES'}]}
Year: 2014
DOI: 10.7907/43RG-P928
In the measurement of the Higgs Boson decaying into two photons the parametrization of an appropriate background model is essential for fitting the Higgs signal mass peak over a continuous background. This diphoton background modeling is crucial in the statistical process of calculating exclusion limits and the significance of observations in comparison to a background-only hypothesis. It is therefore ideal to obtain knowledge of the physical shape for the background mass distribution as the use of an improper function can lead to biases in the observed limits. Using an Information-Theoretic (I-T) approach for valid inference we apply Akaike Information Criterion (AIC) as a measure of the separation for a fitting model from the data. We then implement a multi-model inference ranking method to build a fit-model that closest represents the Standard Model background in 2013 diphoton data recorded by the Compact Muon Solenoid (CMS) experiment at the Large Hadron Collider (LHC). Potential applications and extensions of this model-selection technique are discussed with reference to CMS detector performance measurements as well as in potential physics analyses at future detectors.https://thesis.library.caltech.edu/id/eprint/8520Search for Higgs Boson Production Beyond the Standard Model Using the Razor Kinematic Variables in pp Collisions at √s=8 TeV and Optimization of Higgs Boson Identification Using a Quantum Annealer
https://resolver.caltech.edu/CaltechTHESIS:05052015-155442785
Authors: {'items': [{'email': 'armott@gmail.com', 'id': 'Mott-Alexander-Robert', 'name': {'family': 'Mott', 'given': 'Alexander Robert'}, 'show_email': 'NO'}]}
Year: 2015
DOI: 10.7907/Z9XK8CHR
<p>In the first part of this thesis we search for beyond the Standard Model physics through the search for anomalous production of the Higgs boson using the razor kinematic variables. We search for anomalous Higgs boson production using proton-proton collisions at center of mass energy √s=8 TeV collected by the Compact Muon Solenoid experiment at the Large Hadron Collider corresponding to an integrated luminosity of 19.8 fb<sup>-1</sup>.</p>
<p>In the second part we present a novel method for using a quantum annealer to train a classifier to recognize events containing a Higgs boson decaying to two photons. We train that classifier using simulated proton-proton collisions at √s=8 TeV producing either a Standard Model Higgs boson decaying to two photons or a non-resonant Standard Model process that produces a two photon final state.</p>
<p>The production mechanisms of the Higgs boson are precisely predicted by the Standard Model based on its association with the mechanism of electroweak symmetry breaking. We measure the yield of Higgs bosons decaying to two photons in kinematic regions predicted to have very little contribution from a Standard Model Higgs boson and search for an excess of events, which would be evidence of either non-standard production or non-standard properties of the Higgs boson. We divide the events into disjoint categories based on kinematic properties and the presence of additional b-quarks produced in the collisions. In each of these disjoint categories, we use the razor kinematic variables to characterize events with topological configurations incompatible with typical configurations found from standard model production of the Higgs boson.</p>
<p>We observe an excess of events with di-photon invariant mass compatible with the Higgs boson mass and localized in a small region of the razor plane. We observe 5 events with a predicted background of 0.54 ± 0.28, which observation has a p-value of 10<sup>-3</sup> and a local significance of 3.35σ. This background prediction comes from 0.48 predicted non-resonant background events and 0.07 predicted SM higgs boson events. We proceed to investigate the properties of this excess, finding that it provides a very compelling peak in the di-photon invariant mass distribution and is physically separated in the razor plane from predicted background. Using another method of measuring the background and significance of the excess, we find a 2.5σ deviation from the Standard Model hypothesis over a broader range of the razor plane.</p>
<p>In the second part of the thesis we transform the problem of training a classifier to distinguish events with a Higgs boson decaying to two photons from events with other sources of photon pairs into the Hamiltonian of a spin system, the ground state of which is the best classifier. We then use a quantum annealer to find the ground state of this Hamiltonian and train the classifier. We find that we are able to do this successfully in less than 400 annealing runs for a problem of median difficulty at the largest problem size considered. The networks trained in this manner exhibit good classification performance, competitive with the more complicated machine learning techniques, and are highly resistant to overtraining. We also find that the nature of the training gives access to additional solutions that can be used to improve the classification performance by up to 1.2% in some regions.</p>https://thesis.library.caltech.edu/id/eprint/8845New Physics Models in the Diphoton Final State at CMS
https://resolver.caltech.edu/CaltechTHESIS:01152016-102117113
Authors: {'items': [{'email': 'sezata@gmail.com', 'id': 'Wang-Ann-Miao', 'name': {'family': 'Wang', 'given': 'Ann Miao'}, 'show_email': 'NO'}]}
Year: 2015
DOI: 10.7907/Z9CC0XMN
Since the discovery of the Higgs boson at the LHC, its use as a probe to search for beyond the standard model physics, such as supersymmetry, has become important, as seen in a recent search by the CMS experiment using razor variables in the diphoton final state. Motivated by this search, this thesis examines the LHC discovery potential of a SUSY scenario involving bottom squark pair production with a Higgs boson in the final state. We design and implement a software-based trigger using the razor variables for the CMS experiment to record events with a bottom quark-antiquark pair from a Higgs boson. We characterize the full range of signatures at the LHC from this Higgs-aware SUSY scenario and demonstrate the sensitivity of the CMS data to this model.
https://thesis.library.caltech.edu/id/eprint/9535Extraction of CP Properties of the H(125) Boson Discovered in Proton-Proton Collisions at sqrt(s) = 7 and 8 TeV with the CMS Detector at the LHC
https://resolver.caltech.edu/CaltechTHESIS:12012014-145657884
Authors: {'items': [{'email': 'chen.yi.first@gmail.com', 'id': 'Chen-Yi', 'name': {'family': 'Chen', 'given': 'Yi'}, 'show_email': 'YES'}]}
Year: 2015
DOI: 10.7907/VT37-HT20
In this thesis we build a novel analysis framework to perform the direct extraction of all possible effective Higgs boson couplings to the neutral electroweak gauge bosons in the H → ZZ<sup>(*)</sup> → 4l channel also referred to as the golden channel. We use analytic expressions of the full decay differential cross sections for the H → VV' → 4l process, and the dominant irreducible standard model qq ̄ → 4l background where 4l = 2e2μ,4e,4μ. Detector effects are included through an explicit convolution of these analytic expressions with transfer functions that model the detector responses as well as acceptance and efficiency effects. Using the full set of decay observables, we construct an unbinned 8-dimensional detector level likelihood function which is con- tinuous in the effective couplings, and includes systematics. All potential anomalous couplings of HVV' where V = Z,γ are considered, allowing for general CP even/odd admixtures and any possible phases. We measure the CP-odd mixing between the tree-level HZZ coupling and higher order CP-odd couplings to be compatible with zero, and in the range [−0.40, 0.43], and the mixing between HZZ tree-level coupling and higher order CP -even coupling to be in the ranges [−0.66, −0.57] ∪ [−0.15, 1.00]; namely compatible with a standard model Higgs. We discuss the expected precision in determining the various HVV' couplings in future LHC runs. A powerful and at first glance surprising prediction of the analysis is that with 100-400 fb<sup>-1</sup>, the golden channel will be able to start probing the couplings of the Higgs boson to diphotons in the 4l channel. We discuss the implications and further optimization of the methods for the next LHC runs.https://thesis.library.caltech.edu/id/eprint/8733Searches for New Physics at the Compact Muon Solenoid Experiment and Precision Timing Calorimetry
https://resolver.caltech.edu/CaltechTHESIS:05192017-195511866
Authors: {'items': [{'email': 'cristian.morgoth@gmail.com', 'id': 'Peña-Herrera-Cristián', 'name': {'family': 'Peña Herrera', 'given': 'Cristián'}, 'orcid': '0000-0002-4500-7930', 'show_email': 'NO'}]}
Year: 2017
DOI: 10.7907/Z9MK69Z6
<p>In this thesis, we present several searches for beyond the standard model physics in proton-proton collisions recorded by the Compact Muon Solenoid Experiment at center-of-mass energy of 8 and 13\TeV. We search for particle dark matter in events with two or more jets and missing transverse momentum at $\sqrt{s} = 8$ TeV, in this search we use the razor variables to discriminate signal from background events and thus improve the overall sensitivity of the analysis. We observe agreement between the observation and the background estimation. The interpretation of the results is carried out in the context of an effective field theory that couples the standard model quarks to the dark matter candidate. A search for anomalous production of Higgs bosons using 15.3 $\mathrm{fb}^{-1}$ of proton-proton collisions at $\sqrt{s} = 13$ TeV is also presented; this search selects events with a Higgs boson in association with jets, where the Higgs candidate decays into two photons. We also employ the razor variables ($\mathrm{M_{R}}, \mathrm{R}^{2}$) to discriminate signal from background. We observe an excess of events in one of the search bins with relatively high values of $\mathrm{M_{R}}$ and $\mathrm{R}^{2}$. The interpretation of this analysis is pair production of bottom squarks in the context of supersymmetry, this model is also presented in one of the appendices of this thesis. In the other appendix of this thesis, we present a search for new phenomena in high-mass diphoton events using 12.9 $\mathrm{fb}^{-1}$ of proton-proton collisions at $\sqrt{s} = 13$ TeV. This search observed a significant excess (3.4 standard deviations, local) with 2015 data at a diphoton invariant mass of 750 GeV, equivalent to $\approx$ 20\% of the current dataset. By repeating the search with the larger dataset collected in 2016, we found that the aforementioned excess has been greatly disfavored. Additionally, in order to confirm the robustness and correctness of the data analysis techniques used in this search, we have carried out a second -- completely independent -- analysis, which confirms the absence of an excess at a diphoton invariant mass of 750 GeV.</p>
<p>We also present detector research and developments studies of electromagnetic calorimeters equipped with precision timing capabilities. We present several calorimeter prototypes that were tested at the Fermilab Test Beam Facility. These prototypes include LYSO-based calorimeters, tungsten-LYSO "shashlik" sampling calorimeters, micro-channel-plate sampling calorimeters, and silicon-based sampling calorimeters. The results of these studies indicate that time resolutions of the order of $\sim$ 30 ps are readily available when measuring electromagnetic showers. A discussion about the applications of precision timing in high energy physics experiments is also presented, with a particular interest in pileup rejection in the context of the high-luminosity upgrade of the Large Hadron Collider foreseen to start in 2025.</p>https://thesis.library.caltech.edu/id/eprint/10189Naturalness Confronts Nature: Searches for Supersymmetry with the CMS Detector in pp Collisions at √s = 8 and 13 TeV
https://resolver.caltech.edu/CaltechTHESIS:08232016-112522351
Authors: {'items': [{'email': 'javier.m.g.duarte@gmail.com', 'id': 'Duarte-Javier-Mauricio-Gonzalez', 'name': {'family': 'Duarte', 'given': 'Javier Mauricio Gonzalez'}, 'orcid': '0000-0002-5076-7096', 'show_email': 'NO'}]}
Year: 2017
DOI: 10.7907/Z9GX48JV
In this thesis, we present two inclusive searches for supersymmetric particles at 8 and 13 TeV using the razor variables and guided by the principle of naturalness. We build a framework to explore the natural supersymmetry parameter space of gluino and top squark masses and branching ratios, which is a unique attempt to cover this parameter space in a more complete way than ever before using LHC data. With this approach, the production of top squarks and gluinos are excluded below 700 GeV and 1.6 TeV, respectively, independent of the branching ratios, constituting one of the tightest constraints on natural supersymmetry from the LHC. Motivated by the need to mitigate the effects of multiple interactions per bunch crossing (pileup), an essential feature of present and future hadron colliders, in this thesis we also study the precision timing capabilities of a LYSO-based sampling calorimeter, and achieve a time resolution of 30 ps in electron test beam measurements. The achieved resolution corresponds to the precision needed to significantly reduce the inclusion of pileup particles in the reconstruction of the event of interest. This study is foundational in building an R and D program on precision timing for the high-luminosity LHC and other future hadron colliders. We also propose alternative simplified models to study Higgs-plus-jets events at the LHC, and reinterpret an excess observed at 8 TeV in the context of these models. Finally, we discuss a search for narrow resonances in the dijet mass spectrum at 13 TeV using the data-scouting technique at CMS, which records a smaller event format to increase the maximum recordable rate. For the benchmark models with a vector or axial-vector mediator that couples to quarks and dark matter particles, the dijet search excludes mediator masses from 0.5 TeV up to 2.7 TeV largely independent of the dark matter particle mass, which constitutes a larger exclusion than traditional mono-X searches at the LHC. In the plane of the dark matter-nucleon interaction cross section versus dark matter mass, the dijet search is also more sensitive than direct detection experiments for spin-dependent cross sections.https://thesis.library.caltech.edu/id/eprint/9904Inclusive Searches for Supersymmetry at √s = 13 TeV Using Razor Kinematic Variables, and Data Scouting Using the CMS Trigger System
https://resolver.caltech.edu/CaltechTHESIS:05162018-153827333
Authors: {'items': [{'email': 'dustinanderson111@gmail.com', 'id': 'Anderson-Dustin-James', 'name': {'family': 'Anderson', 'given': 'Dustin James'}, 'orcid': '0000-0001-8173-3182', 'show_email': 'NO'}]}
Year: 2018
DOI: 10.7907/BHGN-VY70
<p>We present two searches for supersymmetric particles using proton-proton collision data collected by the CMS experiment at √s = 13 TeV. The searches use razor kinematic variables for signal discrimination and target the pair production of heavy gluinos and squarks in <i>R</i>-parity conserving supersymmetry. The first search is performed on 2.3 fb<sup>−1</sup> of data collected by CMS in 2015. Two complete, independent background predictions are made: one based on fits using a parameterized functional form, and the other based on Monte Carlo simulation corrected via control samples in data. The second search is an expanded version of the first search, and is performed using the Monte Carlo-based background prediction method on 35.9 fb<sup>−1</sup> of data collected in 2016. Both searches obtain results compatible with standard model background expectations. The null results are interpreted as limits on the masses and cross sections of gluinos, squarks, and higgsinos in the context of simplified models of supersymmetry. We discuss the outlook for the fit-based search strategy and explore how the technique of gaussian process regression may be useful as a tool to combat the challenges of this analysis methodology.</p>
<p>We also describe a new paradigm for trigger-level collider data analysis, which we refer to as data scouting. In this paradigm, searches for new physics are performed using event information reconstructed within the experiment’s trigger software. This circumvents traditional event rate constraints, such as disk space and the latency of offline reconstruction. We provide details on the implementation of a framework for data scouting in the CMS High-Level Trigger system and its successful use in Run II of the LHC. We discuss the impact of scouting on the physics program of CMS and demonstrate that it enables searches for new physics that would not otherwise be possible due to trigger constraints, such as hadronic resonance searches at low mass and searches for leptonic decays of dark photons.</p>https://thesis.library.caltech.edu/id/eprint/10902Search for SUSY with Delayed Photons at the Compact Muon Solenoid
https://resolver.caltech.edu/CaltechTHESIS:05222018-110636485
Authors: {'items': [{'email': 'gillian.kopp@gmail.com', 'id': 'Kopp-Gillian-Baron', 'name': {'family': 'Kopp', 'given': 'Gillian Baron'}, 'orcid': '0000-0001-8160-0208', 'show_email': 'NO'}]}
Year: 2018
DOI: 10.7907/G4D7-G385
The Compact Muon Solenoid (CMS) experiment records data from proton-proton (pp) and heavy ion (Pb-Pb and Pb-p) collisions at the Large Hadron Collider (LHC) to search for physics beyond the Standard Model, test theories of supersymmetry (SUSY), and measure properties of known particles with higher precision. In 2025, the LHC will be upgraded to the High Luminosity LHC (HL-LHC), where the luminosity will be increased by a factor of 10. This will increase the number of pile-up collisions to 140-200 events per proton-proton bunch crossing, compared to the current 40 events per crossing (where each bunch crossing occurs every 25 ns). In order to fully exploit the sensitivity of the CMS experiment, the current detectors must be upgraded to mitigate the effects of the large number of pileup interactions expected in collisions at the HL-LHC. New capabilities, such as precision timing measurements in calorimetric devices and minimum ionizing detectors, have been shown to effectively mitigate the effects due to pileup, and are expected to benefit the overall physics reach of the experiment. In addition to mitigating pileup and increasing the detector capabilities, precision timing is beneficial in the search for particles beyond the Standard Model. A simulation of a benchmark long lived neutralino SUSY search is presented, and it is shown that the generator particle flight times can be faithfully reconstructed using the detector-level information. Identification algorithms for the SUSY model have been significantly improved with the use of a Boosted Decision Tree, and it is demonstrated that this algorithm has many benefits as compared to cut based IDs. With use of the BDT for the long lived neutralino SUSY model, the background rejection is increased significantly, with constant signal acceptance of 53.6%. This is an improvement in the significance of the signal selection by a factor of 2.38. Further improvement is seen with the inclusion of detector timing information in the BDT – with this contributing ≈25% of the information used in signal event identification. We thus demonstrate that with the BDT, the SUSY neutralino search can be performed with increased signal identification significance, and the searches’ sensitivity is expected to improve with the time resolution attained by the upgraded CMS calorimeter. https://thesis.library.caltech.edu/id/eprint/10924Quantum Computing for Machine Learning and Physics Simulation
https://resolver.caltech.edu/CaltechTHESIS:09272022-143825909
Authors: {'items': [{'email': 'azlokapa@mit.edu', 'id': 'Zlokapa-Alexander', 'name': {'family': 'Zlokapa', 'given': 'Alexander'}, 'orcid': '0000-0002-4153-8646', 'show_email': 'NO'}]}
Year: 2021
DOI: 10.7907/q75q-zm20
<p>Quantum computing is widely thought to provide exponential speedups over classical algorithms for a variety of computational tasks. In classical computing, methods in artificial intelligence such as neural networks and adversarial learning have enabled drastic improvements in state-of-the-art performance for a variety of tasks. We consider the intersection of quantum computing with machine learning, including the quantum algorithms for deep learning on classical datasets, quantum adversarial learning for quantum states, and variational quantum machine learning for improved physics simulation.</p>
<p>We consider a standard deep neural network architecture and show that conditions amenable to trainability by gradient descent coincide with those necessary for an efficient quantum algorithm. Considering the neural network in the infinite-width limit using the neural tangent kernel formalism, we propose a quantum algorithm to train the neural network with vanishing error as the training dataset size increases. Under a sparse approximation of the neural tangent kernel, the training time scales logarithmically with the number of training examples, providing the first known exponential quantum speedup for feedforward neural networks. Related approximations to the neural tangent kernel are discussed, with numerical studies showing successful convergence beyond the proven regime. Our work suggests the applicability of the quantum computing to additional neural network architectures and common datasets such as MNIST, as well as kernel methods beyond the neural tangent kernel.</p>
<p>Generative adversarial networks (GANs) are one of the most widely adopted machine learning methods for data generation. We propose an entangling quantum GAN (EQ-GAN) that overcomes some limitations of previously proposed quantum GANs. EQ-GAN guarantees the convergence to a Nash equilibrium under minimax optimization of the discriminator and generator circuits by performing entangling operations between both the generator output and true quantum data. We show that EQ-GAN has additional robustness against coherent errors and demonstrate the effectiveness of EQ-GAN experimentally in a Google Sycamore superconducting quantum processor. By adversarially learning efficient representations of quantum states, we prepare an approximate quantum random access memory and demonstrate its use in applications including the training of near-term quantum neural networks.</p>
<p>With quantum computers providing a natural platform for physics simulation, we investigate the use of variational quantum circuits to simulate many-body systems with high fidelity in the near future. In particular, recent work shows that teleportation caused by introducing a weak coupling between two entangled SYK models is dual to a particle traversing an AdS-Schwarzschild wormhole, providing a mechanism to probe quantum gravity theories in the lab. To simulate such a system, we propose the process of compressed Trotterization to improve the fidelity of time evolution on noisy devices. The task of learning approximate time evolution circuits is shown to have a favorable training landscape, and numerical experiments demonstrate its relevance to simulating other many-body systems such as a Fermi-Hubbard model. For the SYK model in particular, we demonstrate the construction of a low-rank approximation that favors a shallower Trotterization. Finally, classical simulations of finite-N SYK models suggest that teleportation via a traversable wormhole instead of random unitary scrambling is achievable with O(20) qubits, providing further indication that such quantum gravity experiments may realizable with near-term quantum hardware.</p>https://thesis.library.caltech.edu/id/eprint/15035Searches for Nonresonant Higgs Boson Pair Production and Long-Lived Particles at the LHC and Machine-Learning Solutions for the High-Luminosity LHC Era
https://resolver.caltech.edu/CaltechTHESIS:10012021-210141406
Authors: {'items': [{'email': 'thongnguyen1292@gmail.com', 'id': 'Nguyen-Thong-Quang', 'name': {'family': 'Nguyen', 'given': 'Thong Quang'}, 'orcid': '0000-0003-3954-5131', 'show_email': 'NO'}]}
Year: 2022
DOI: 10.7907/knfz-q495
<p>This thesis presents two physics analyses using 137 fb<sup>−1</sup> proton-proton collision data collected by the CMS experiment at √s = 13 TeV, along with a series of machine-learning solutions to extend the physics program at the LHC and to address the computational challenges in the High-Luminosity LHC era. The first analysis searches for nonresonant Higgs boson pair production in final states with two photons and two bottom quarks, with no significant deviation from the background-only hypothesis observed. The observed (expected) upper limit on the product of the Higgs boson pair production cross section and branching fraction into bb̅γγ is 0.67 (0.45) fb, corresponding to 7.7 (5.2) times the Standard Model prediction. The modifier of the Higgs trilinear self-coupling is constrained within the range -3.3 < κ<sub>λ</sub> < 8.5. The modifier for coupling between a pair of Higgs bosons and a pair of vector bosons, along with the 2-dimensional constraint of the modifiers of Higgs self-coupling and Yukawa coupling, are also reported. A graph-based algorithm to identify boosted H → bb̅ jets to improve future Higgs search is presented. The second analysis searches for long-lived supersymmetry particles decaying to photons and gravitinos in the context of gauge-mediated supersymmetry breaking model. Results are presented in terms of 95% confidence level expected exclusion limits on the masses and proper decay lengths of the neutralino, which exceed the limits from the previous searches by up to 100 GeV for the neutralino mass and by five times for the neutralino proper decay length. A strategy for model-independent new physics searches is presented with an anomaly trigger based on unsupervised learning algorithms that can be deployed in both the high-level trigger and the Level-1 trigger in CMS. Three other machine-learning solutions are presented to address the computational challenges in the HL-LHC era: a layer based on multi-modal deep neural networks that can reduce the false-positive events selected by the trigger by over one order of magnitude while retaining 99% of signal events, a full-event simulation algorithm based on recurrent generative adversarial networks that has potential to replace traditional simulation method while being five orders of magnitude faster, and a fast simulation algorithm for specific analyses based on encoder-decoder architecture that would result in about an order-of-magnitude reduction in computing and storage requirements for the collision simulation workflow.</p>https://thesis.library.caltech.edu/id/eprint/14380New Physics Tools for Discovery, a New Era of Timing Detector, and Lepton Flavor Universality Test at CMS
https://resolver.caltech.edu/CaltechTHESIS:09132022-001004681
Authors: {'items': [{'email': 'olmo.cerri@gmail.com', 'id': 'Cerri-Olmo', 'name': {'family': 'Cerri', 'given': 'Olmo'}, 'orcid': '0000-0002-2191-0666', 'show_email': 'NO'}]}
Year: 2023
DOI: 10.7907/bqsn-sp82
<p>The field of particle and fundamental physics finds itself now in a peculiar situation. The established Standard Model accurately predicts most of the observations, but several compelling reasons motivate a need for an extension of the current theory. In this thesis, I focus my research on facing the current situation of the field in a diversified threefold manner.</p>
<p>First, I develop methods based on physics-driven machine learning algorithms, with a particular focus on developing a model-independent tagger for unexpected events using artificial neural networks. This study shows how model-independent new physics triggers, possibly trained on real data, can select a low rate stream of events able to explore new physics processes up to a 10-100 pb cross section and can create a special dataset of rare unexpected events. Other important results from this body of work include the first application of the proposed anomaly detection strategy to real data, the use of graph neural networks to improve current pileup mitigation algorithms, the development of jet taggers based on the interaction network, and analysis-specific fast simulation.</p>
<p>Second, I focus on the methodological and hardware development of the MIP Timing Layer that is expected to upgrade CMS in preparation for HL-LHC. My seminal study demonstrates the possibility of using time-of-flight information to perform particle identification, which has a significant impact on heavy stable charged particle searches. This work introduces how to measure time-of-flight at CMS, a strategy for particle identification, and an algorithm to locate vertices in space and time. I also participated in the sensor testing and test beam operation. In particular, I conducted a study about the design and prototype of the detector modules' thermal behavior that shows how different geometries could lead to cooling differences of a few K.</p>
<p>Last, I direct my attention towards CMS's first lepton flavor universality tests with B meson decays. Using a dataset acquired thanks to a custom design trigger, I independently develop the measurement of the R(D*) ratio, a parameter whose tensions between the predictions and observation have drawn remarkable attentions. I oversaw the complete mature state of the analysis, from the Monte Carlo simulation to the fitting procedure. Further collaboration-wide efforts are still required, but I demonstrate the expected sensitivity of about 15% using an Asimov dataset.</p>https://thesis.library.caltech.edu/id/eprint/15019A Deep Dive into the Connections Between the Renormalization Group and Deep Learning in the Ising Model
https://resolver.caltech.edu/CaltechTHESIS:05302023-084739008
Authors: {'items': [{'email': 'kelsietaylor137@gmail.com', 'id': 'Taylor-Kelsie-Reed', 'name': {'family': 'Taylor', 'given': 'Kelsie'}, 'orcid': '0009-0001-7510-2306', 'show_email': 'YES'}]}
Year: 2023
DOI: 10.7907/ztpg-z092
The renormalization group (RG) is an essential technique in statistical physics and quantum field theory, which considers scale-invariant properties of physical theories and how these theories’ parameters change with scaling. Deep learning is a powerful computational technique that uses multi-layered neural networks to solve a myriad of complicated problems. Previous research suggests the possibility that unsupervised deep learning may be a form of RG flow, by being a layer-by-layer coarse graining of the original data. We examined this connection on a more rigorous basis for the simple example of Kadanoff block renormalization of the 2D nearest-neighbor Ising model, with our deep learning accomplished via Restricted Boltzmann Machines (RBMs). We developed extensive renormalization techniques for the 1D and 2D Ising model to provide a baseline for comparison. For the 1D Ising model, we successfully used Adam optimization on a correlation length loss function to learn the group flow; yielding results consistent with the analytical model for infinite N. For the 2D Ising model, we successfully generated Ising model samples using the Wolff algorithm, and performed the group flow using a quasi-deterministic method, validating these results by calculating the critical exponent \nu. We then examined RBM learning of the Ising model layer by layer, finding a blocking structure in the learning that is qualitatively similar to RG. Lastly, we directly compared the weights of each layer from the learning to Ising spin renormalization, but found quantitative inconsistencies for the simple case of nearest-neighbor Ising models.https://thesis.library.caltech.edu/id/eprint/15230Rare Higgs Processes at CMS and Precision Timing Detector Studies for HL-LHC CMS Upgrade
https://resolver.caltech.edu/CaltechTHESIS:09252022-195535688
Authors: {'items': [{'email': 'irene.230594@gmail.com', 'id': 'Dutta-Irene', 'name': {'family': 'Dutta', 'given': 'Irene'}, 'orcid': '0000-0003-0953-4503', 'show_email': 'NO'}]}
Year: 2023
DOI: 10.7907/tmt4-nq20
<p>This thesis describes the search for two rare Higgs processes. The first analysis describes the CMS Run 2 search for <it>H → µµ</it> decays, with 137.3 fb<sup>-1</sup> of data at <it>√s</it> = 13 TeV. The analysis targeted four different Higgs production modes: the gluon fusion (ggH), the vector boson fusion (VBF), the Higgs-strahlung process (VH), and the production in association with a pair of top quarks (ttH). Each category used a dedicated machine learning based classifier to separate the signal from the background processes. A combined fit from all these categories saw a slight excess in the data corresponding to 3.0 standard deviations at <it>M<sub>H</sub></it> = 125.38 GeV, and gave the first evidence for the Higgs boson decay to second-generation fermions. The best-fit signal strength and the corresponding 68% CL interval was found to be µ^ =1.19 <sup>+0.41</sup><sub>-0.39</sub> (stat) <sup>+0.17</sup><sub>-0.16</sub> (syst) at <it>M<sub>H</sub></it> = 125.38 GeV. </p>
<p>The second analysis describes the CMS Run 2 search for <it>HH → bb̅bb̅</it> with highly boosted Higgs bosons. This analysis used a dedicated jet identification algorithm based on graph neural networks (ParticleNet) to identify boosted H → bb jets. This search targeted the gluon fusion and the vector boson fusion HH production modes, and put constraints on the allowed values of the various Higgs couplings as: κ<sub>λ</sub> ∈ [-9.9,16.9] when κ<sub>ν</sub> = 1, κ<sub>2ν</sub>= 1; κ<sub>ν</sub> ∈ [-1.17,-0.79] ∪ [0.81,1.18] when κ<sub>λ</sub> = 1, κ<sub>2ν</sub> = 1; κ<sub>2ν</sub> ∈ [0.62,1.41] when κ<sub>λ</sub> = 1, κ<sub>ν</sub> = 1. A scenario with κ<sub>2ν</sub> = 0 was excluded with a significance of 6.3 standard deviations for the first time, when other H couplings are fixed to their SM values. The combined observed (expected) 95% upper limit on the HH production cross section was found to be 9.9 (5.1) x SM. </p>
<p> Finally, this thesis also discusses the planned MIP Timing Detector (MTD) upgrade for CMS at the HL-LHC. The MTD will be a time-of-flight (TOF) detector, designed to provide a precision timing information for charged particles using SiPMs + LYSO scintillating crystals, with a time resolution of ~30 ps. This thesis describes several R&D tests that have been performed for characterizing the sensor properties (time resolution, light yield, etc.) and optimizing the sensor design geometry. This thesis also contains a description of mock test setups for cooling the sensors, since it is known to be an effective way of mitigating the increased dark current rates in the sensors due to radiation damage. </p>https://thesis.library.caltech.edu/id/eprint/15034Search for Supersymmetry Using Higgs Boson to Diphoton Decays and Search for Long-Lived Particles Using Out-of-Time Trackless Jets at √s = 13 TeV
https://resolver.caltech.edu/CaltechTHESIS:05302023-234037855
Authors: {'items': [{'email': 'mjiajing@gmail.com', 'id': 'Mao-Jiajing', 'name': {'family': 'Mao', 'given': 'Jiajing'}, 'orcid': '0009-0002-8988-9987', 'show_email': 'NO'}]}
Year: 2023
DOI: 10.7907/pwpe-ef70
<p>The thesis describes two searches conducted at the Large Hadron Collider with a center-of-mass energy of 13 TeV, using proton-proton collison data collected by the CMS experiment. The supersymmetry (SUSY) search focus on the production of at least one Higgs boson that decays into two photons in the decay chains of pair-produced SUSY particles. The data used has an integrated luminosity of 77.5 fb⁻¹. The events are classified into different search regions based on charged leptons, Higgs boson candidates, and kinematic variables to make them sensitive to different SUSY scenarios. The results reveal no statistically significant excess of events compared to the standard model predictions. The searches exclude bottom squark pair production for bottom squark masses below 510 GeV and a lightest SUSY particle mass of 1 GeV. The wino-like chargino-neutralino production in gauge-mediated SUSY breaking (GMSB) is excluded for chargino and neutralino masses below 235 GeV, with a gravitino mass of 1 GeV. Furthermore, the higgsino-like chargino-neutralino production in GMSB, where the neutralino decays exclusively to a Higgs boson and a gravitino, is excluded for neutralino masses below 290 GeV. The thesis also reports a search for long-lived particles that decay in the outer regions of the CMS silicon tracker or in the calorimeters. The search uses data with an integrated luminosity of 138 fb⁻¹. The identification of long-lived particle decays utilizes a novel technique that combines nearly trackless and out-of-time jet information into a deep neural network discriminator. The results are interpreted using a simplified GMSB model of chargino-neutralino production, where the neutralino is the next-to-lightest supersymmetric particle that decays to a gravitino and either a Higgs or Z boson. The search achieves the highest sensitivity for neutralino proper decay lengths of approximately 0.5 meters and excludes masses up to 1.18 TeV at a 95\% confidence level. This search represents the most stringent constraint to date in the mass range from the kinematic limit imposed by the Higgs boson mass up to 1.8 TeV.</p>https://thesis.library.caltech.edu/id/eprint/15237