-
Deep Learning Model Deployment in Multiple Cloud Providers: an Exploratory Study Using Low Computing Power Environments
Authors:
Elayne Lemos,
Rodrigo Oliveira,
Jairson Rodrigues,
Rosalvo F. Oliveira Neto
Abstract:
The deployment of Machine Learning models at cloud have grown by tech companies. Hardware requirements are higher when these models involve Deep Learning (DL) techniques and the cloud providers' costs may be a barrier. We explore deploying DL models using for experiments the GECToR model, a DL solution for Grammatical Error Correction, across three of the major cloud platforms (AWS, Google Cloud,…
▽ More
The deployment of Machine Learning models at cloud have grown by tech companies. Hardware requirements are higher when these models involve Deep Learning (DL) techniques and the cloud providers' costs may be a barrier. We explore deploying DL models using for experiments the GECToR model, a DL solution for Grammatical Error Correction, across three of the major cloud platforms (AWS, Google Cloud, Azure). We evaluate real-time latency, hardware usage and cost at each cloud provider by 7 execution environments with 10 experiments reproduced. We found that while GPUs excel in performance, they had an average cost 300% higher than solutions without GPU. Our analysis also identifies that processor cache size is crucial for cost-effective CPU deployments, enabling over 50% of cost reduction compared to GPUs. This study demonstrates the feasibility and affordability of cloud-based DL inference solutions without GPUs, benefiting resource-constrained users like startups.
△ Less
Submitted 31 March, 2025;
originally announced March 2025.
-
Search for reactor-produced millicharged particles with Skipper-CCDs at the CONNIE and Atucha-II experiments
Authors:
Alexis A. Aguilar-Arevalo,
Nicolas Avalos,
Pablo Bellino,
Xavier Bertou,
Carla Bonifazi,
Ana Botti,
Mariano Cababié,
Gustavo Cancelo,
Brenda A. Cervantes-Vergara,
Claudio Chavez,
Fernando Chierchie,
David Delgado,
Eliana Depaoli,
Juan Carlos D'Olivo,
João dos Anjos,
Juan Estrada,
Guillermo Fernandez Moroni,
Aldo R. Fernandes Neto,
Richard Ford,
Ben Kilminster,
Kevin Kuk,
Andrew Lathrop,
Patrick Lemos,
Herman P. Lima Jr.,
Martin Makler
, et al. (15 additional authors not shown)
Abstract:
Millicharged particles, proposed by various extensions of the standard model, can be created in pairs by high-energy photons within nuclear reactors and can interact electromagnetically with electrons in matter. Recently, the existence of a plasmon peak in the interaction cross-section with silicon in the eV range was highlighted as a promising approach to enhance low-energy sensitivities. The CON…
▽ More
Millicharged particles, proposed by various extensions of the standard model, can be created in pairs by high-energy photons within nuclear reactors and can interact electromagnetically with electrons in matter. Recently, the existence of a plasmon peak in the interaction cross-section with silicon in the eV range was highlighted as a promising approach to enhance low-energy sensitivities. The CONNIE and Atucha-II reactor neutrino experiments utilize Skipper-CCD sensors, which enable the detection of interactions in the eV range. We present world-leading limits on the charge of millicharged particles within a mass range spanning six orders of magnitude, derived through a comprehensive analysis and the combination of data from both experiments.
△ Less
Submitted 5 November, 2024; v1 submitted 25 May, 2024;
originally announced May 2024.
-
Searches for CEνNS and Physics beyond the Standard Model using Skipper-CCDs at CONNIE
Authors:
Alexis A. Aguilar-Arevalo,
Nicolas Avalos,
Xavier Bertou,
Carla Bonifazi,
Gustavo Cancelo,
Brenda A. Cervantes-Vergara,
Claudio Chavez,
Fernando Chierchie,
Gustavo Coelho Corrêa,
Juan Carlos D'Olivo,
João dos Anjos,
Juan Estrada,
Guillermo Fernandez Moroni,
Aldo R. Fernandes Neto,
Richard Ford,
Ben Kilminster,
Kevin Kuk,
Andrew Lathrop,
Patrick Lemos,
Herman P. Lima Jr.,
Martin Makler,
Katherine Maslova,
Franciole Marinho,
Jorge Molina,
Irina Nasteva
, et al. (9 additional authors not shown)
Abstract:
The Coherent Neutrino-Nucleus Interaction Experiment (CONNIE) aims to detect the coherent scattering (CE$ν$NS) of reactor antineutrinos off silicon nuclei using thick fully-depleted high-resistivity silicon CCDs. Two Skipper-CCD sensors with sub-electron readout noise capability were installed at the experiment next to the Angra-2 reactor in 2021, making CONNIE the first experiment to employ Skipp…
▽ More
The Coherent Neutrino-Nucleus Interaction Experiment (CONNIE) aims to detect the coherent scattering (CE$ν$NS) of reactor antineutrinos off silicon nuclei using thick fully-depleted high-resistivity silicon CCDs. Two Skipper-CCD sensors with sub-electron readout noise capability were installed at the experiment next to the Angra-2 reactor in 2021, making CONNIE the first experiment to employ Skipper-CCDs for reactor neutrino detection. We report on the performance of the Skipper-CCDs, the new data processing and data quality selection techniques and the event selection for CE$ν$NS interactions, which enable CONNIE to reach a record low detection threshold of 15 eV. The data were collected over 300 days in 2021-2022 and correspond to exposures of 14.9 g-days with the reactor-on and 3.5 g-days with the reactor-off. The difference between the reactor-on and off event rates shows no excess and yields upper limits at 95% confidence level for the neutrino interaction rates comparable with previous CONNIE limits from standard CCDs and higher exposures. Searches for new neutrino interactions beyond the Standard Model were performed, yielding an improvement on the previous CONNIE limit on a simplified model with light vector mediators. A first dark matter (DM) search by diurnal modulation was performed by CONNIE and the results represent the best limits on the DM-electron scattering cross-section, obtained by a surface-level experiment. These promising results, obtained using a very small-mass sensor, illustrate the potential of Skipper-CCDs to probe rare neutrino interactions and motivate the plans to increase the detector mass in the near future.
△ Less
Submitted 23 March, 2024;
originally announced March 2024.
-
Exact Markovian evolution of quantum systems with several degrees of freedom : Phase space representations
Authors:
Aldo R. Fernandes Neto,
Alfredo M. Ozorio de Almeida,
Olivier Brodier
Abstract:
The exact solution of the Lindblad equation with a quadratic Hamiltonian and linear coupling operators was derived within the chord representation, that is, for the Fourier transform of the Wigner function, also known as the characteristic function. It is here generalized for several degrees of freedom, so as to provide an explicit expression for the reduced density operator of any subsystem, as w…
▽ More
The exact solution of the Lindblad equation with a quadratic Hamiltonian and linear coupling operators was derived within the chord representation, that is, for the Fourier transform of the Wigner function, also known as the characteristic function. It is here generalized for several degrees of freedom, so as to provide an explicit expression for the reduced density operator of any subsystem, as well as moments expressed as derivatives of this evolving chord function. The Wigner function is then the convolution of its straightforward classical evolution with a widening multidimensional Gaussian window, eventually ensuring its positivity. Futher on, positivity also holds for the Glauber-Sundarshan P function, which guarantees separability of the components. In the context of several degrees of freedom, a full dissipation matrix is defined, whose trace is equal to twice the previously derived dissipation coefficient. This governs the rate at which the phase space volume of the argument of the Wigner function contracts, while that of the chord function expands. Examples of Markovian evolution of a triatomic molecule and of an array of harmonic oscillators are discussed.
△ Less
Submitted 3 October, 2023; v1 submitted 3 August, 2022;
originally announced August 2022.
-
Search for coherent elastic neutrino-nucleus scattering at a nuclear reactor with CONNIE 2019 data
Authors:
CONNIE collaboration,
Alexis Aguilar-Arevalo,
Javier Bernal,
Xavier Bertou,
Carla Bonifazi,
Gustavo Cancelo,
Victor G. P. B. de Carvalho,
Brenda A. Cervantes-Vergara,
Claudio Chavez,
Gustavo Coelho Corrêa,
Juan C. D'Olivo,
João C. dos Anjos,
Juan Estrada,
Aldo R. Fernandes Neto,
Guillermo Fernandez Moroni,
Ana Foguel,
Richard Ford,
Julián Gasanego Barbuscio,
Juan Gonzalez Cuevas,
Susana Hernandez,
Federico Izraelevitch,
Ben Kilminster,
Kevin Kuk,
Herman P. Lima Jr,
Martin Makler
, et al. (11 additional authors not shown)
Abstract:
The Coherent Neutrino-Nucleus Interaction Experiment (CONNIE) is taking data at the Angra 2 nuclear reactor with the aim of detecting the coherent elastic scattering of reactor antineutrinos with silicon nuclei using charge-coupled devices (CCDs). In 2019 the experiment operated with a hardware binning applied to the readout stage, leading to lower levels of readout noise and improving the detecti…
▽ More
The Coherent Neutrino-Nucleus Interaction Experiment (CONNIE) is taking data at the Angra 2 nuclear reactor with the aim of detecting the coherent elastic scattering of reactor antineutrinos with silicon nuclei using charge-coupled devices (CCDs). In 2019 the experiment operated with a hardware binning applied to the readout stage, leading to lower levels of readout noise and improving the detection threshold down to 50 eV. The results of the analysis of 2019 data are reported here, corresponding to the detector array of 8 CCDs with a fiducial mass of 36.2 g and a total exposure of 2.2 kg-days. The difference between the reactor-on and reactor-off spectra shows no excess at low energies and yields upper limits at 95% confidence level for the neutrino interaction rates. In the lowest-energy range, 50-180 eV, the expected limit stands at 34 (39) times the standard model prediction, while the observed limit is 66 (75) times the standard model prediction with Sarkis (Chavarria) quenching factors.
△ Less
Submitted 6 April, 2022; v1 submitted 25 October, 2021;
originally announced October 2021.
-
Search for light mediators in the low-energy data of the CONNIE reactor neutrino experiment
Authors:
Alexis Aguilar-Arevalo,
Xavier Bertou,
Carla Bonifazi,
Gustavo Cancelo,
Brenda A. Cervantes-Vergara,
Claudio Chavez,
Juan C. D'Olivo,
João C. dos Anjos,
Juan Estrada,
Aldo R. Fernandes Neto,
Guillermo Fernandez-Moroni,
Ana Foguel,
Richard Ford,
Federico Izraelevitch,
Ben Kilminster,
H. P. Lima Jr,
Martin Makler,
Jorge Molina,
Philipe Mota,
Irina Nasteva,
Eduardo Paolini,
Carlos Romero,
Youssef Sarkis,
Miguel Sofo-Haro,
Javier Tiffenberg
, et al. (1 additional authors not shown)
Abstract:
The CONNIE experiment is located at a distance of 30 m from the core of a commercial nuclear reactor, and has collected a 3.7 kg-day exposure using a CCD detector array sensitive to an $\sim$1 keV threshold for the study of coherent neutrino-nucleus elastic scattering. Here we demonstrate the potential of this low-energy neutrino experiment as a probe for physics Beyond the Standard Model, by usin…
▽ More
The CONNIE experiment is located at a distance of 30 m from the core of a commercial nuclear reactor, and has collected a 3.7 kg-day exposure using a CCD detector array sensitive to an $\sim$1 keV threshold for the study of coherent neutrino-nucleus elastic scattering. Here we demonstrate the potential of this low-energy neutrino experiment as a probe for physics Beyond the Standard Model, by using the recently published results to constrain two simplified extensions of the Standard Model with light mediators. We compare the new limits with those obtained for the same models using neutrinos from the Spallation Neutron Source. Our new constraints represent the best limits for these simplified models among the experiments searching for CE$ν$NS for a light vector mediator with mass $M_{Z^{\prime}}<$ 10 MeV, and for a light scalar mediator with mass $M_φ<$ 30 MeV. These results constitute the first use of the CONNIE data as a probe for physics Beyond the Standard Model.
△ Less
Submitted 29 March, 2020; v1 submitted 10 October, 2019;
originally announced October 2019.
-
Exploring low-energy neutrino physics with the Coherent Neutrino Nucleus Interaction Experiment (CONNIE)
Authors:
Alexis Aguilar-Arevalo,
Xavier Bertou,
Carla Bonifazi,
Gustavo Cancelo,
Alejandro Castañeda,
Brenda Cervantes Vergara,
Claudio Chavez,
Juan C. D'Olivo,
João C. dos Anjos,
Juan Estrada,
Aldo R. Fernandes Neto,
Guillermo Fernandez Moroni,
Ana Foguel,
Richard Ford,
Juan Gonzalez Cuevas,
Pamela Hernández,
Susana Hernandez,
Federico Izraelevitch,
Alexander R. Kavner,
Ben Kilminster,
Kevin Kuk,
H. P. Lima Jr,
Martin Makler,
Jorge Molina,
Philipe Mota
, et al. (8 additional authors not shown)
Abstract:
The Coherent Neutrino-Nucleus Interaction Experiment (CONNIE) uses low-noise fully depleted charge-coupled devices (CCDs) with the goal of measuring low-energy recoils from coherent elastic scattering (CE$ν$NS) of reactor antineutrinos with silicon nuclei and testing nonstandard neutrino interactions (NSI). We report here the first results of the detector array deployed in 2016, considering an act…
▽ More
The Coherent Neutrino-Nucleus Interaction Experiment (CONNIE) uses low-noise fully depleted charge-coupled devices (CCDs) with the goal of measuring low-energy recoils from coherent elastic scattering (CE$ν$NS) of reactor antineutrinos with silicon nuclei and testing nonstandard neutrino interactions (NSI). We report here the first results of the detector array deployed in 2016, considering an active mass 47.6 g (8 CCDs), which is operating at a distance of 30 m from the core of the Angra 2 nuclear reactor, with a thermal power of 3.8 GW. A search for neutrino events is performed by comparing data collected with reactor on (2.1 kg-day) and reactor off (1.6 kg-day). The results show no excess in the reactor-on data, reaching the world record sensitivity down to recoil energies of about 1 keV (0.1 keV electron-equivalent). A 95% confidence level limit for new physics is established at an event rate of 40 times the one expected from the standard model at this energy scale. The results presented here provide a new window to low-energy neutrino physics, allowing one to explore for the first time the energies accessible through the low threshold of CCDs. They will lead to new constrains on NSI from the CE$ν$NS of antineutrinos from nuclear reactors.
△ Less
Submitted 10 October, 2019; v1 submitted 5 June, 2019;
originally announced June 2019.
-
A Music Classification Model based on Metric Learning and Feature Extraction from MP3 Audio Files
Authors:
Angelo C. Mendes da Silva,
Mauricio A. Nunes,
Raul Fonseca Neto
Abstract:
The development of models for learning music similarity and feature extraction from audio media files is an increasingly important task for the entertainment industry. This work proposes a novel music classification model based on metric learning and feature extraction from MP3 audio files. The metric learning process considers the learning of a set of parameterized distances employing a structure…
▽ More
The development of models for learning music similarity and feature extraction from audio media files is an increasingly important task for the entertainment industry. This work proposes a novel music classification model based on metric learning and feature extraction from MP3 audio files. The metric learning process considers the learning of a set of parameterized distances employing a structured prediction approach from a set of MP3 audio files containing several music genres. The main objective of this work is to make possible learning a personalized metric for each customer. To extract the acoustic information we use the Mel-Frequency Cepstral Coefficient (MFCC) and make a dimensionality reduction with the use of Principal Components Analysis. We attest the model validity performing a set of experiments and comparing the training and testing results with baseline algorithms, such as K-means and Soft Margin Linear Support Vector Machine (SVM). Experiments show promising results and encourage the future development of an online version of the learning model.
△ Less
Submitted 17 September, 2019; v1 submitted 29 May, 2019;
originally announced May 2019.
-
Optimal Categorical Attribute Transformation for Granularity Change in Relational Databases for Binary Decision Problems in Educational Data Mining
Authors:
Paulo J. L. Adeodato,
Fábio C. Pereira,
Rosalvo F. Oliveira Neto
Abstract:
This paper presents an approach for transforming data granularity in hierarchical databases for binary decision problems by applying regression to categorical attributes at the lower grain levels. Attributes from a lower hierarchy entity in the relational database have their information content optimized through regression on the categories histogram trained on a small exclusive labelled sample, i…
▽ More
This paper presents an approach for transforming data granularity in hierarchical databases for binary decision problems by applying regression to categorical attributes at the lower grain levels. Attributes from a lower hierarchy entity in the relational database have their information content optimized through regression on the categories histogram trained on a small exclusive labelled sample, instead of the usual mode category of the distribution. The paper validates the approach on a binary decision task for assessing the quality of secondary schools focusing on how logistic regression transforms the students and teachers attributes into school attributes. Experiments were carried out on Brazilian schools public datasets via 10-fold cross-validation comparison of the ranking score produced also by logistic regression. The proposed approach achieved higher performance than the usual distribution mode transformation and equal to the expert weighing approach measured by the maximum Kolmogorov-Smirnov distance and the area under the ROC curve at 0.01 significance level.
△ Less
Submitted 28 February, 2017;
originally announced February 2017.
