Search

Your search keyword '"Wolfe, Cameron R"' showing total 480 results
Start Over Author "Wolfe, Cameron R"

Refine your results

more »

more »

more »

more »
480 results on '"Wolfe, Cameron R"'

1. Better Schedules for Low Precision Training of Deep Neural Networks

Author

Wolfe, Cameron R. and Kyrillidis, Anastasios

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence, I.2.6, I.2.10, I.4.0
Abstract

Low precision training can significantly reduce the computational overhead of training deep neural networks (DNNs). Though many such techniques exist, cyclic precision training (CPT), which dynamically adjusts precision throughout training according to a cyclic schedule, achieves particularly impressive improvements in training efficiency, while actually improving DNN performance. Existing CPT implementations take common learning rate schedules (e.g., cyclical cosine schedules) and use them for low precision training without adequate comparisons to alternative scheduling options. We define a diverse suite of CPT schedules and analyze their performance across a variety of DNN training regimes, some of which are unexplored in the low precision training literature (e.g., node classification with graph neural networks). From these experiments, we discover alternative CPT schedules that offer further improvements in training efficiency and model performance, as well as derive a set of best practices for choosing CPT schedules. Going further, we find that a correlation exists between model performance and training cost, and that changing the underlying CPT schedule can control the tradeoff between these two variables. To explain the direct correlation between model performance and training cost, we draw a connection between quantized training and critical learning periods, suggesting that aggressive quantization is a form of learning impairment that can permanently damage model performance., Comment: 20 pages, 8 figures, 1 table, ACML 2023

Published
2024
Full Text
View/download PDF

3. Evaluating Demographic Representation in Clinical Trials: Use of the Adaptive Coronavirus Disease 2019 Treatment Trial (ACTT) as a Test Case.

Author

Ortega-Villa, Ana M, Hynes, Noreen A, Levine, Corri B, Yang, Katherine, Wiley, Zanthia, Jilg, Nikolaus, Wang, Jing, Whitaker, Jennifer A, Colombo, Christopher J, Nayak, Seema U, Kim, Hannah Jang, Iovine, Nicole M, Ince, Dilek, Cohen, Stuart H, Langer, Adam J, Wortham, Jonathan M, Atmar, Robert L, El Sahly, Hana M, Jain, Mamta K, Mehta, Aneesh K, Wolfe, Cameron R, Gomez, Carlos A, Beresnev, Tatiana, Mularski, Richard A, Paules, Catharine I, Kalil, Andre C, Branche, Angela R, Luetkemeyer, Annie, Zingman, Barry S, Voell, Jocelyn, Whitaker, Michael, Harkins, Michelle S, Davey, Richard T, Grossberg, Robert, George, Sarah L, Tapson, Victor, Short, William R, Ghazaryan, Varduhi, Benson, Constance A, Dodd, Lori E, Sweeney, Daniel A, and Tomashek, Kay M

Subjects
ACTT, COVID-19 clinical trials, representation evaluation, Clinical Research, Infectious Diseases, Prevention, Clinical Trials and Supportive Activities, Good Health and Well Being
Abstract

BackgroundClinical trials initiated during emerging infectious disease outbreaks must quickly enroll participants to identify treatments to reduce morbidity and mortality. This may be at odds with enrolling a representative study population, especially when the population affected is undefined.MethodsWe evaluated the utility of the Centers for Disease Control and Prevention's COVID-19-Associated Hospitalization Surveillance Network (COVID-NET), the COVID-19 Case Surveillance System (CCSS), and 2020 United States (US) Census data to determine demographic representation in the 4 stages of the Adaptive COVID-19 Treatment Trial (ACTT). We compared the cumulative proportion of participants by sex, race, ethnicity, and age enrolled at US ACTT sites, with respective 95% confidence intervals, to the reference data in forest plots.ResultsUS ACTT sites enrolled 3509 adults hospitalized with COVID-19. When compared with COVID-NET, ACTT enrolled a similar or higher proportion of Hispanic/Latino and White participants depending on the stage, and a similar proportion of African American participants in all stages. In contrast, ACTT enrolled a higher proportion of these groups when compared with US Census and CCSS. The proportion of participants aged ≥65 years was either similar or lower than COVID-NET and higher than CCSS and the US Census. The proportion of females enrolled in ACTT was lower than the proportion of females in the reference datasets.ConclusionsAlthough surveillance data of hospitalized cases may not be available early in an outbreak, they are a better comparator than US Census data and surveillance of all cases, which may not reflect the population affected and at higher risk of severe disease.

Published
2023

4. Baricitinib Treatment of Coronavirus Disease 2019 Is Associated With a Reduction in Secondary Infections

Author

Sweeney, Daniel A, Tuyishimire, Bonifride, Ahuja, Neera, Beigel, John H, Beresnev, Tatiana, Cantos, Valeria D, Castro, Jose G, Cohen, Stuart H, Cross, Kaitlyn, Dodd, Lori E, Erdmann, Nathan, Fung, Monica, Ghazaryan, Varduhi, George, Sarah L, Grimes, Kevin A, Hynes, Noreen A, Julian, Kathleen G, Kandiah, Sheetal, Kim, Hannah Jang, Levine, Corri B, Lindholm, David A, Lye, David C, Maves, Ryan C, Oh, Myoung-Don, Paules, Catharine, Rapaka, Rekha R, Short, Willam R, Tomashek, Kay M, Wolfe, Cameron R, and Kalil, Andre C

Subjects
Biomedical and Clinical Sciences, Clinical Sciences, Infectious Diseases, Clinical Trials and Supportive Activities, Clinical Research, Evaluation of treatments and therapeutic interventions, 6.1 Pharmaceuticals, Good Health and Well Being, baricitinib, COVID-19, secondary infections, Clinical sciences, Medical microbiology
Abstract

We performed a secondary analysis of the National Institutes of Health-sponsored Adaptive COVID-19 Treatment Trial (ACTT-2) randomized controlled trial and found that baricitinib was associated with a 50% reduction in secondary infections after controlling for baseline and postrandomization patient characteristics. This finding provides a novel mechanism of benefit for baricitinib and supports the safety profile of this immunomodulator for the treatment of coronavirus disease 2019.

Published
2023

5. Cold Start Streaming Learning for Deep Networks

Author

Wolfe, Cameron R. and Kyrillidis, Anastasios

Subjects
Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition, Mathematics - Optimization and Control, 68T07, I.2.6, I.2.10, I.4.0
Abstract

The ability to dynamically adapt neural networks to newly-available data without performance deterioration would revolutionize deep learning applications. Streaming learning (i.e., learning from one data example at a time) has the potential to enable such real-time adaptation, but current approaches i) freeze a majority of network parameters during streaming and ii) are dependent upon offline, base initialization procedures over large subsets of data, which damages performance and limits applicability. To mitigate these shortcomings, we propose Cold Start Streaming Learning (CSSL), a simple, end-to-end approach for streaming learning with deep networks that uses a combination of replay and data augmentation to avoid catastrophic forgetting. Because CSSL updates all model parameters during streaming, the algorithm is capable of beginning streaming from a random initialization, making base initialization optional. Going further, the algorithm's simplicity allows theoretical convergence guarantees to be derived using analysis of the Neural Tangent Random Feature (NTRF). In experiments, we find that CSSL outperforms existing baselines for streaming learning in experiments on CIFAR100, ImageNet, and Core50 datasets. Additionally, we propose a novel multi-task streaming learning setting and show that CSSL performs favorably in this domain. Put simply, CSSL performs well and demonstrates that the complicated, multi-step training pipelines adopted by most streaming methodologies can be replaced with a simple, end-to-end learning approach without sacrificing performance., Comment: 52 pages, 7 figures, pre-print

Published
2022

6. Temporal Improvements in COVID-19 Outcomes for Hospitalized Adults: A Post Hoc Observational Study of Remdesivir Group Participants in the Adaptive COVID-19 Treatment Trial

Author

Potter, Gail E, Bonnett, Tyler, Rubenstein, Kevin, Lindholm, David A, Rapaka, Rekha R, Doernberg, Sarah B, Lye, David C, Mularski, Richard A, Hynes, Noreen A, Kline, Susan, Paules, Catharine I, Wolfe, Cameron R, Frank, Maria G, Rouphael, Nadine G, Deye, Gregory A, Sweeney, Daniel A, Colombo, Rhonda E, Davey, Richard T, Mehta, Aneesh K, Whitaker, Jennifer A, Castro, Jose G, Amin, Alpesh N, Colombo, Christopher J, Levine, Corri B, Jain, Mamta K, Maves, Ryan C, Marconi, Vincent C, Grossberg, Robert, Hozayen, Sameh, Burgess, Timothy H, Atmar, Robert L, Ganesan, Anuradha, Gomez, Carlos A, Benson, Constance A, de Castilla, Diego Lopez, Ahuja, Neera, George, Sarah L, Nayak, Seema U, Cohen, Stuart H, Lalani, Tahaniyat, Short, William R, Erdmann, Nathaniel, Tomashek, Kay M, and Tebas, Pablo

Subjects
Biomedical and Clinical Sciences, Clinical Sciences, Prevention, Clinical Research, Rehabilitation, Complementary and Integrative Health, Clinical Trials and Supportive Activities, 6.1 Pharmaceuticals, Evaluation of treatments and therapeutic interventions, Good Health and Well Being, Adult, Humans, Antiviral Agents, Clinical Trials, Phase III as Topic, Dexamethasone, Double-Blind Method, Randomized Controlled Trials as Topic, Treatment Outcome, COVID-19 Drug Treatment, Medical and Health Sciences, General & Internal Medicine, Clinical sciences
Abstract

BackgroundThe COVID-19 standard of care (SOC) evolved rapidly during 2020 and 2021, but its cumulative effect over time is unclear.ObjectiveTo evaluate whether recovery and mortality improved as SOC evolved, using data from ACTT (Adaptive COVID-19 Treatment Trial).DesignACTT is a series of phase 3, randomized, double-blind, placebo-controlled trials that evaluated COVID-19 therapeutics from February 2020 through May 2021. ACTT-1 compared remdesivir plus SOC to placebo plus SOC, and in ACTT-2 and ACTT-3, remdesivir plus SOC was the control group. This post hoc analysis compared recovery and mortality between these comparable sequential cohorts of patients who received remdesivir plus SOC, adjusting for baseline characteristics with propensity score weighting. The analysis was repeated for participants in ACTT-3 and ACTT-4 who received remdesivir plus dexamethasone plus SOC. Trends in SOC that could explain outcome improvements were analyzed. (ClinicalTrials.gov: NCT04280705 [ACTT-1], NCT04401579 [ACTT-2], NCT04492475 [ACTT-3], and NCT04640168 [ACTT-4]).Setting94 hospitals in 10 countries (86% U.S. participants).ParticipantsAdults hospitalized with COVID-19.InterventionSOC.Measurements28-day mortality and recovery.ResultsAlthough outcomes were better in ACTT-2 than in ACTT-1, adjusted hazard ratios (HRs) were close to 1 (HR for recovery, 1.04 [95% CI, 0.92 to 1.17]; HR for mortality, 0.90 [CI, 0.56 to 1.40]). Comparable patients were less likely to be intubated in ACTT-2 than in ACTT-1 (odds ratio, 0.75 [CI, 0.53 to 0.97]), and hydroxychloroquine use decreased. Outcomes improved from ACTT-2 to ACTT-3 (HR for recovery, 1.43 [CI, 1.24 to 1.64]; HR for mortality, 0.45 [CI, 0.21 to 0.97]). Potential explanatory factors (SOC trends, case surges, and variant trends) were similar between ACTT-2 and ACTT-3, except for increased dexamethasone use (11% to 77%). Outcomes were similar in ACTT-3 and ACTT-4. Antibiotic use decreased gradually across all stages.LimitationUnmeasured confounding.ConclusionChanges in patient composition explained improved outcomes from ACTT-1 to ACTT-2 but not from ACTT-2 to ACTT-3, suggesting improved SOC. These results support excluding nonconcurrent controls from analysis of platform trials in rapidly changing therapeutic areas.Primary funding sourceNational Institute of Allergy and Infectious Diseases.

Published
2022

7. PipeGCN: Efficient Full-Graph Training of Graph Convolutional Networks with Pipelined Feature Communication

Author

Wan, Cheng, Li, Youjie, Wolfe, Cameron R., Kyrillidis, Anastasios, Kim, Nam Sung, and Lin, Yingyan

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract

Graph Convolutional Networks (GCNs) is the state-of-the-art method for learning graph-structured data, and training large-scale GCNs requires distributed training across multiple accelerators such that each accelerator is able to hold a partitioned subgraph. However, distributed GCN training incurs prohibitive overhead of communicating node features and feature gradients among partitions for every GCN layer during each training iteration, limiting the achievable training efficiency and model scalability. To this end, we propose PipeGCN, a simple yet effective scheme that hides the communication overhead by pipelining inter-partition communication with intra-partition computation. It is non-trivial to pipeline for efficient GCN training, as communicated node features/gradients will become stale and thus can harm the convergence, negating the pipeline benefit. Notably, little is known regarding the convergence rate of GCN training with both stale features and stale feature gradients. This work not only provides a theoretical convergence analysis but also finds the convergence rate of PipeGCN to be close to that of the vanilla distributed GCN training without any staleness. Furthermore, we develop a smoothing method to further improve PipeGCN's convergence. Extensive experiments show that PipeGCN can largely boost the training throughput (1.7x~28.5x) while achieving the same accuracy as its vanilla counterpart and existing full-graph training methods. The code is available at https://github.com/RICE-EIC/PipeGCN., Comment: ICLR 2022

Published
2022

8. Baricitinib versus dexamethasone for adults hospitalised with COVID-19 (ACTT-4): a randomised, double-blind, double placebo-controlled trial.

Author

Wolfe, Cameron R, Tomashek, Kay M, Patterson, Thomas F, Gomez, Carlos A, Marconi, Vincent C, Jain, Mamta K, Yang, Otto O, Paules, Catharine I, Palacios, Guillermo M Ruiz, Grossberg, Robert, Harkins, Michelle S, Mularski, Richard A, Erdmann, Nathaniel, Sandkovsky, Uriel, Almasri, Eyad, Pineda, Justino Regalado, Dretler, Alexandra W, de Castilla, Diego Lopez, Branche, Angela R, Park, Pauline K, Mehta, Aneesh K, Short, William R, McLellan, Susan LF, Kline, Susan, Iovine, Nicole M, El Sahly, Hana M, Doernberg, Sarah B, Oh, Myoung-Don, Huprikar, Nikhil, Hohmann, Elizabeth, Kelley, Colleen F, Holodniy, Mark, Kim, Eu Suk, Sweeney, Daniel A, Finberg, Robert W, Grimes, Kevin A, Maves, Ryan C, Ko, Emily R, Engemann, John J, Taylor, Barbara S, Ponce, Philip O, Larson, LuAnn, Melendez, Dante Paolo, Seibert, Allan M, Rouphael, Nadine G, Strebe, Joslyn, Clark, Jesse L, Julian, Kathleen G, de Leon, Alfredo Ponce, Cardoso, Anabela, de Bono, Stephanie, Atmar, Robert L, Ganesan, Anuradha, Ferreira, Jennifer L, Green, Michelle, Makowski, Mat, Bonnett, Tyler, Beresnev, Tatiana, Ghazaryan, Varduhi, Dempsey, Walla, Nayak, Seema U, Dodd, Lori E, Beigel, John H, Kalil, Andre C, and ACTT-4 Study Group

Subjects
ACTT-4 Study Group, Humans, Oxygen, Sulfonamides, Azetidines, Pyrazoles, Purines, Dexamethasone, Treatment Outcome, Double-Blind Method, Adolescent, Adult, Middle Aged, Female, Male, SARS-CoV-2, COVID-19 Drug Treatment, Clinical Trials and Supportive Activities, Lung, Clinical Research, Prevention, 6.1 Pharmaceuticals, Evaluation of treatments and therapeutic interventions, Good Health and Well Being, Clinical Sciences, Public Health and Health Services, Other Medical and Health Sciences
Abstract

BackgroundBaricitinib and dexamethasone have randomised trials supporting their use for the treatment of patients with COVID-19. We assessed the combination of baricitinib plus remdesivir versus dexamethasone plus remdesivir in preventing progression to mechanical ventilation or death in hospitalised patients with COVID-19.MethodsIn this randomised, double-blind, double placebo-controlled trial, patients were enrolled at 67 trial sites in the USA (60 sites), South Korea (two sites), Mexico (two sites), Singapore (two sites), and Japan (one site). Hospitalised adults (≥18 years) with COVID-19 who required supplemental oxygen administered by low-flow (≤15 L/min), high-flow (>15 L/min), or non-invasive mechanical ventilation modalities who met the study eligibility criteria (male or non-pregnant female adults ≥18 years old with laboratory-confirmed SARS-CoV-2 infection) were enrolled in the study. Patients were randomly assigned (1:1) to receive either baricitinib, remdesivir, and placebo, or dexamethasone, remdesivir, and placebo using a permuted block design. Randomisation was stratified by study site and baseline ordinal score at enrolment. All patients received remdesivir (≤10 days) and either baricitinib (or matching oral placebo) for a maximum of 14 days or dexamethasone (or matching intravenous placebo) for a maximum of 10 days. The primary outcome was the difference in mechanical ventilation-free survival by day 29 between the two treatment groups in the modified intention-to-treat population. Safety analyses were done in the as-treated population, comprising all participants who received one dose of the study drug. The trial is registered with ClinicalTrials.gov, NCT04640168.FindingsBetween Dec 1, 2020, and April 13, 2021, 1047 patients were assessed for eligibility. 1010 patients were enrolled and randomly assigned, 516 (51%) to baricitinib plus remdesivir plus placebo and 494 (49%) to dexamethasone plus remdesivir plus placebo. The mean age of the patients was 58·3 years (SD 14·0) and 590 (58%) of 1010 patients were male. 588 (58%) of 1010 patients were White, 188 (19%) were Black, 70 (7%) were Asian, and 18 (2%) were American Indian or Alaska Native. 347 (34%) of 1010 patients were Hispanic or Latino. Mechanical ventilation-free survival by day 29 was similar between the study groups (Kaplan-Meier estimates of 87·0% [95% CI 83·7 to 89·6] in the baricitinib plus remdesivir plus placebo group and 87·6% [84·2 to 90·3] in the dexamethasone plus remdesivir plus placebo group; risk difference 0·6 [95% CI -3·6 to 4·8]; p=0·91). The odds ratio for improved status in the dexamethasone plus remdesivir plus placebo group compared with the baricitinib plus remdesivir plus placebo group was 1·01 (95% CI 0·80 to 1·27). At least one adverse event occurred in 149 (30%) of 503 patients in the baricitinib plus remdesivir plus placebo group and 179 (37%) of 482 patients in the dexamethasone plus remdesivir plus placebo group (risk difference 7·5% [1·6 to 13·3]; p=0·014). 21 (4%) of 503 patients in the baricitinib plus remdesivir plus placebo group had at least one treatment-related adverse event versus 49 (10%) of 482 patients in the dexamethasone plus remdesivir plus placebo group (risk difference 6·0% [2·8 to 9·3]; p=0·00041). Severe or life-threatening grade 3 or 4 adverse events occurred in 143 (28%) of 503 patients in the baricitinib plus remdesivir plus placebo group and 174 (36%) of 482 patients in the dexamethasone plus remdesivir plus placebo group (risk difference 7·7% [1·8 to 13·4]; p=0·012).InterpretationIn hospitalised patients with COVID-19 requiring supplemental oxygen by low-flow, high-flow, or non-invasive ventilation, baricitinib plus remdesivir and dexamethasone plus remdesivir resulted in similar mechanical ventilation-free survival by day 29, but dexamethasone was associated with significantly more adverse events, treatment-related adverse events, and severe or life-threatening adverse events. A more individually tailored choice of immunomodulation now appears possible, where side-effect profile, ease of administration, cost, and patient comorbidities can all be considered.FundingNational Institute of Allergy and Infectious Diseases.

Published
2022

9. i-SpaSP: Structured Neural Pruning via Sparse Signal Recovery

Author

Wolfe, Cameron R. and Kyrillidis, Anastasios

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Mathematics - Optimization and Control, Statistics - Machine Learning, 68T07, I.2.6, I.2.10, I.4.0
Abstract

We propose a novel, structured pruning algorithm for neural networks -- the iterative, Sparse Structured Pruning algorithm, dubbed as i-SpaSP. Inspired by ideas from sparse signal recovery, i-SpaSP operates by iteratively identifying a larger set of important parameter groups (e.g., filters or neurons) within a network that contribute most to the residual between pruned and dense network output, then thresholding these groups based on a smaller, pre-defined pruning ratio. For both two-layer and multi-layer network architectures with ReLU activations, we show the error induced by pruning with i-SpaSP decays polynomially, where the degree of this polynomial becomes arbitrarily large based on the sparsity of the dense network's hidden representations. In our experiments, i-SpaSP is evaluated across a variety of datasets (i.e., MNIST, ImageNet, and XNLI) and architectures (i.e., feed forward networks, ResNet34, MobileNetV2, and BERT), where it is shown to discover high-performing sub-networks and improve upon the pruning efficiency of provable baseline methodologies by several orders of magnitude. Put simply, i-SpaSP is easy to implement with automatic differentiation, achieves strong empirical results, comes with theoretical convergence guarantees, and is efficient, thus distinguishing itself as one of the few computationally efficient, practical, and provable pruning algorithms., Comment: 29 pages, 4 figures, 4th Annual Conference on Learning for Dynamics and Control

Published
2021

10. National Landscape of Human Immunodeficiency Virus-Positive Deceased Organ Donors in the United States.

Author

Werbel, William A, Brown, Diane M, Kusemiju, Oyinkansola T, Doby, Brianna L, Seaman, Shanti M, Redd, Andrew D, Eby, Yolanda, Fernandez, Reinaldo E, Desai, Niraj M, Miller, Jernelle, Bismut, Gilad A, Kirby, Charles S, Schmidt, Haley A, Clarke, William A, Seisa, Michael, Petropoulos, Christos J, Quinn, Thomas C, Florman, Sander S, Huprikar, Shirish, Rana, Meenakshi M, Friedman-Moraco, Rachel J, Mehta, Aneesh K, Stock, Peter G, Price, Jennifer C, Stosor, Valentina, Mehta, Shikha G, Gilbert, Alexander J, Elias, Nahel, Morris, Michele I, Mehta, Sapna A, Small, Catherine B, Haidar, Ghady, Malinis, Maricar, Husson, Jennifer S, Pereira, Marcus R, Gupta, Gaurav, Hand, Jonathan, Kirchner, Varvara A, Agarwal, Avinash, Aslam, Saima, Blumberg, Emily A, Wolfe, Cameron R, Myer, Kevin, Wood, R Patrick, Neidlinger, Nikole, Strell, Sara, Shuck, Marion, Wilkins, Harry, Wadsworth, Matthew, Motter, Jennifer D, Odim, Jonah, Segev, Dorry L, Durand, Christine M, Tobian, Aaron AR, Piquant, Dominque, Link, Katherine, Hemmersbach-Miller, Marion, Pearson, Thomas, Turgeon, Nicole, Lyon, G Marshall, Kitchens, William, Huckaby, Jeryl, Lasseter, A Francie, Elbein, Rivka, Roberson, April, Ferry, Elizabeth, Klock, Ethan, Cochran, Willa V, Morrison, Michelle, Rasmussen, Sarah, Bollinger, Juli, Sugarman, Jeremy, Smith, Angela R, Thomas, Margaret, Coakley, Margaret, Timpone, Joseph, Stucke, Alyssa, Haydel, Brandy, Dieter, Rebecca, Klein, Elizabeth J, Neumann, Henry, Gallon, Lorenzo, Goudy, Leah, Callegari, Michelle, Marrazzo, Ilise, Jackson, Towanda, Pruett, Timothy, Farnsworth, Mary, Locke, Jayme E, Mompoint-Williams, Darnell, Basinger, Katherine, Mekeel, Kristin, Nguyen, Phirum, Kwan, Joanne, Srisengfa, Tab, Chin-Hong, Peter, Rogers, Rodney, Simkins, Jacques, Munoz, Carlos, and Dunn, Ty

Subjects
Pediatric Research Initiative, HIV/AIDS, Infectious Diseases, 6.1 Pharmaceuticals, Evaluation of treatments and therapeutic interventions, Infection, Good Health and Well Being, Anti-Retroviral Agents, HIV, HIV Infections, HIV Seropositivity, Humans, Integrases, Prospective Studies, Tissue Donors, United States, Viral Load, transplant, organ donation, drug resistance, HOPE in Action Investigators, Biological Sciences, Medical and Health Sciences, Microbiology
Abstract

BackgroundOrgan transplantation from donors with human immunodeficiency virus (HIV) to recipients with HIV (HIV D+/R+) presents risks of donor-derived infections. Understanding clinical, immunologic, and virologic characteristics of HIV-positive donors is critical for safety.MethodsWe performed a prospective study of donors with HIV-positive and HIV false-positive (FP) test results within the HIV Organ Policy Equity (HOPE) Act in Action studies of HIV D+/R+ transplantation (ClinicalTrials.gov NCT02602262, NCT03500315, and NCT03734393). We compared clinical characteristics in HIV-positive versus FP donors. We measured CD4 T cells, HIV viral load (VL), drug resistance mutations (DRMs), coreceptor tropism, and serum antiretroviral therapy (ART) detection, using mass spectrometry in HIV-positive donors.ResultsBetween March 2016 and March 2020, 92 donors (58 HIV positive, 34 FP), representing 98.9% of all US HOPE donors during this period, donated 177 organs (131 kidneys and 46 livers). Each year the number of donors increased. The prevalence of hepatitis B (16% vs 0%), syphilis (16% vs 0%), and cytomegalovirus (CMV; 91% vs 58%) was higher in HIV-positive versus FP donors; the prevalences of hepatitis C viremia were similar (2% vs 6%). Most HIV-positive donors (71%) had a known HIV diagnosis, of whom 90% were prescribed ART and 68% had a VL

Published
2022

11. Implementation of a Pooled Surveillance Testing Program for Asymptomatic SARS-CoV-2 Infections on a College Campus -- Duke University, Durham, North Carolina, August 2-October 11, 2020

Author

Centers for Disease Control and Prevention (DHHS/PHS), Denny, Thomas N., Andrews, Laura, Bonsignori, Mattia, Cavanaugh, Kyle, Datto, Michael B., Deckard, Anastasia, DeMarco, Todd, DeNaeyer, Nicole, Epling, Carol A., Gurley, Thaddeus, Haase, Steven B., Hallberg, Chloe, Harer, John, Kneifel, Charles L., Lee, Mark J., Louzao, Raul, Moody, M. Anthony, Moore, Zack, Polage, Christopher R., Puglin, Jamie, Spotts, P. Hunter, Vaughn, John A., and Wolfe, Cameron R.

Abstract

In fall 2020, Duke University's COVID-19 prevention strategy included risk reduction behaviors, frequent testing using pooled SARS-CoV-2 polymerase chain reaction testing, and contact tracing. Among 10,265 students who received testing 68,913 times, 84 had positive results. One half of infections were asymptomatic, and some had high viral loads. SARS-CoV-2 transmission was limited in this congregate setting by integration of prevention strategies that included identification of asymptomatic infections through frequent testing. Pooled testing reduced the need for resources while allowing high throughput with high sensitivity and rapid turnaround of results.

Published
2020

12. Remdesivir for the prevention of invasive mechanical ventilation or death in COVID-19 - A post-hoc analysis of the Adaptive COVID-19 Treatment Trial-1 Cohort Data

Author

Paules, Catharine I, Gallagher, Shannon K, Rapaka, Rekha R, Davey, Richard T, Doernberg, Sarah B, Grossberg, Robert, Hynes, Noreen A, Ponce, Philip, Short, William R, Voell, Jocelyn, Wang, Jing, Yang, Otto O, Wolfe, Cameron R, Lye, David C, Dodd, Lori E, and Benson, Constance A

Subjects
Clinical Research, Prevention, Clinical Trials and Supportive Activities, Lung, Neurodegenerative, Good Health and Well Being, Adenosine Monophosphate, Alanine, Antiviral Agents, Clinical Trials as Topic, Humans, Respiration, Artificial, SARS-CoV-2, COVID-19 Drug Treatment, COVID-19, remdesivir, ACTT-1, Biological Sciences, Medical and Health Sciences, Microbiology
Abstract

This post hoc analysis of the Adaptive Coronavirus Disease 2019 (COVID-19) Treatment Trial-1 (ACTT-1) shows a treatment effect of remdesivir (RDV) on progression to invasive mechanical ventilation (IMV) or death. Additionally, we create a risk profile that better predicts progression than baseline oxygen requirement alone. The highest risk group derives the greatest treatment effect from RDV.

Published
2022

13. How much pre-training is enough to discover a good subnetwork?

Author

Wolfe, Cameron R., Liao, Fangshuo, Wang, Qihan, Kim, Junhyung Lyle, and Kyrillidis, Anastasios

Subjects
Statistics - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Machine Learning, Mathematics - Optimization and Control, 68T07, I.2.6, I.2.10, I.4.0
Abstract

Neural network pruning is useful for discovering efficient, high-performing subnetworks within pre-trained, dense network architectures. More often than not, it involves a three-step process -- pre-training, pruning, and re-training -- that is computationally expensive, as the dense model must be fully pre-trained. While previous work has revealed through experiments the relationship between the amount of pre-training and the performance of the pruned network, a theoretical characterization of such dependency is still missing. Aiming to mathematically analyze the amount of dense network pre-training needed for a pruned network to perform well, we discover a simple theoretical bound in the number of gradient descent pre-training iterations on a two-layer, fully-connected network, beyond which pruning via greedy forward selection [61] yields a subnetwork that achieves good training error. Interestingly, this threshold is shown to be logarithmically dependent upon the size of the dataset, meaning that experiments with larger datasets require more pre-training for subnetworks obtained via pruning to perform well. Lastly, we empirically validate our theoretical results on a multi-layer perceptron trained on MNIST., Comment: 29 pages

Published
2021

14. Exceeding the Limits of Visual-Linguistic Multi-Task Learning

Author

Wolfe, Cameron R. and Lundgaard, Keld T.

Subjects
Computer Science - Artificial Intelligence, Computer Science - Computation and Language, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning, 68T07, I.2.6, I.2.7, I.2.10
Abstract

By leveraging large amounts of product data collected across hundreds of live e-commerce websites, we construct 1000 unique classification tasks that share similarly-structured input data, comprised of both text and images. These classification tasks focus on learning the product hierarchy of different e-commerce websites, causing many of them to be correlated. Adopting a multi-modal transformer model, we solve these tasks in unison using multi-task learning (MTL). Extensive experiments are presented over an initial 100-task dataset to reveal best practices for "large-scale MTL" (i.e., MTL with more than 100 tasks). From these experiments, a final, unified methodology is derived, which is composed of both best practices and new proposals such as DyPa, a simple heuristic for automatically allocating task-specific parameters to tasks that could benefit from extra capacity. Using our large-scale MTL methodology, we successfully train a single model across all 1000 tasks in our dataset while using minimal task specific parameters, thereby showing that it is possible to extend several orders of magnitude beyond current efforts in MTL., Comment: 10 pages, 7 figures

Published
2021

15. ResIST: Layer-Wise Decomposition of ResNets for Distributed Training

Author

Dun, Chen, Wolfe, Cameron R., Jermaine, Christopher M., and Kyrillidis, Anastasios

Subjects
Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Distributed, Parallel, and Cluster Computing, Mathematics - Optimization and Control
Abstract

We propose ResIST, a novel distributed training protocol for Residual Networks (ResNets). ResIST randomly decomposes a global ResNet into several shallow sub-ResNets that are trained independently in a distributed manner for several local iterations, before having their updates synchronized and aggregated into the global model. In the next round, new sub-ResNets are randomly generated and the process repeats until convergence. By construction, per iteration, ResIST communicates only a small portion of network parameters to each machine and never uses the full model during training. Thus, ResIST reduces the per-iteration communication, memory, and time requirements of ResNet training to only a fraction of the requirements of full-model training. In comparison to common protocols, like data-parallel training and data-parallel training with local SGD, ResIST yields a decrease in communication and compute requirements, while being competitive with respect to model performance., Comment: 26 pages, 8 figures, pre-print under review

Published
2021

17. Immunomodulators for immunocompromised patients hospitalized for COVID-19: a meta-analysis of randomized controlled trials

Author

Hermine, Olivier, Mariette, Xavier, Ravaud, Philippe, Bureau, Serge, Dougados, Maxime, Resche-Rigon, Matthieu, Tharaux, Pierre-Louis, Tibi, Annick, Azoulay, Elie, Cadranel, Jacques, Emmerich, Joseph, Fartoukh, Muriel, Guidet, Bertrand, Humbert, Marc, Lacombe, Karine, Mahevas, Matthieu, Pene, Frédéric, Porcher, Raphaël, Pourchet-Martinez, Valerie, Schlemmer, Frédéric, Yazdanpanah, Yazdan, Baron, Gabriel, Perrodeau, Elodie, Vanhoye, Damien, Kedzia, Cécile, Demerville, Lauren, Gysembergh-Houal, Anne, Bourgoin, Alexandre, Raked, Nabil, Mameri, Lakhdar, Montlahuc, Claire, Biard, Lucie, Alary, St.phanie, Hamiria, Samir, Bariz, Thinhinane, Semri, Hala, Hai, Dhiaa Meriem, Benafla, Moustafa, Belloul, Mohamed, Vauboin, Pernelle, Flamand, Saskia, Pacheco, Claire, Walter-Petrich, Anouk, Stan, Emilia, Benarab, Souad, Nyanou, Corine, Charreteur, Robin, Dupre, Céline, Cardet, Kévin, Lehmann, Blandine, Baghli, Kamyl, Madelaine, Claire, D'Ortenzio, Eric, Puéchal, Oriane, Semaille, Caroline, Savale, Laurent, Harrois, Anatole, Figueiredo, Samy, Duranteau, Jacques, Anguel, Nadia, Pavot, Arthur, Monnet, Xavier, Richard, Christian, Teboul, Jean-Louis, Durand, Philippe, Tissieres, Pierre, Jevnikar, Mitja, Montani, David, Pavy, Stephan, Nocturne, Gaétane, Bitoun, Samuel, Noel, Nicolas, Lambotte, Olivier, Escaut, Lelia, Jauréguiberry, Stephane, Baudry, Elodie, Verny, Christiane, Lefevre, Edouard, Zaidan, Mohamad, Molinari, Domitille, Leprun, Gaël, Fourreau, Alain, Cylly, Laurent, Grimaldi, Lamiae, Virlouvet, Myriam, Meftali, Ramdane, Fabre, Soléne, Licois, Marion, Mamoune, Asmaa, Boudali, Yacine, Le Tiec, Clotilde, Verstuyft, Céline, Roques, Anne-Marie, Georgin-Lavialle, Sophie, Senet, Patricia, Pialoux, Gilles, Soria, Angele, Parrot, Antoine, François, Helene, Rozensztajn, Nathalie, Blin, Emmanuelle, Choinier, Pascaline, Camuset, Juliette, Rech, Jean-Simon, Canellas, Antony, Rolland-Debord, Camille, Lemarié, Nadege, Belaube, Nicolas, Nadal, Marine, Siguier, Martin, Petit-Hoang, Camille, Chas, Julie, Drouet, Elodie, Lemoine, Matthieu, Phibel, Audrey, Aunay, Lucie, Bertrand, Eliane, Ravato, Sylviane, Vayssettes, Marie, Adda, Anne, Wilpotte, Celine, Thibaut, Pélagie, Fillon, Julie, Debrix, Isabelle, Fellahi, Soraya, Bastard, Jean-Philippe, Lefévre, Guillaume, Gottenberg, Jacques-Eric, Hansmann, Yves, Blanc, Frédéric, Ohlmann-Caillard, Sophie, Castelain, Vincent, Chatelus, Emmanuel, Chatron, Eva, Collange, Olivier, Danion, François, De Blay, Frédéric, Diemunsch, Pierre, Diemunsch, Sophie, Felten, Renaud, Goichot, Bernard, Greigert, Valentin, Guffroy, Aurelien, Heger, Bob, Kaeuffer, Charlotte, Kassegne, Loic, Korganow, Anne Sophie, Le Borgne, Pierrick, Lefebvre, Nicolas, Mertes, Paul-Michel, Noll, Eric, Oberlin, Mathieu, Poindron, Vincent, Pottecher, Julien, Ruch, Yvon, Weill, François, Meyer, Nicolas, Andres, Emmanuel, Demonsant, Eric, Tayebi, Hakim, Nisand, Gabriel, Brin, Stéphane, Sublon, Cédric, Becker, Guillaume, Hutt, Anne, Martin, Tristan, Bayer, Sophie, Metzger, Catherine, Mekinian, Arsene, Abisror, Noémie, Adedjouma, Amir, Bollens, Diane, Bonneton, Marion, Bourcicaux, Nathalie, Bourrier, Anne, Thibault Chiarabiani, Maria Chauchard, Chopin, Doroth.e, Cohen, Jonathan, Devred, Ines, Donadille, Bruno, Fain, Olivier, Hariri, Geoffrey, Jachiet, Vincent, Ingliz, Patrick, Garnier, Marc, Gatfosse, Marc, Ghrenassia, Etienne, Gobert, Delphine, Krause le Garrec, Jessica, Landman, Cecilia, Lavillegrand, Jean Remy, Lefebvre, Benedicte, Mahevas, Thibault, Mazerand, Sandie, Meynard, Jean Luc, Morgand, Marjolaine, Ouaz.ne, Zineb, Pacanowski, Jerome, Riviere, S.bastien, Seksik, Philippe, Sokol, Harry, Soliman, Heithem, Valin, Nadia, Urbina, Thomas, McAvoy, Chloé, Miranda, Maria Pereira, Aratus, Gladys, Berard, Laurence, Simon, Tabassome, Nguyen, Anne Daguenel, Girault, Elise, Mayala-Kanda, Cl.mentine, Antignac, Marie, Leplay, Céline, Arlet, Jean-Benoit, Diehl, Jean-Luc, Bellenfant, Florence, Blanchard, Anne, Buffet, Alexandre, Cholley, Bernard, Fayol, Antoine, Flamarion, Edouard, Godier, Anne, Gorget, Thomas, Hamada, Sophie-Rym, Hauw-Berlemont, Caroline, Hulot, Jean-Sébastien, Lebeaux, David, Livrozet, Marine, Michon, Adrien, Neuschwander, Arthur, Pennet, Marie-Aude, Planquette, Benjamin, Ranque, Brigitte, Sanchez, Olivier, Volle, Geoffroy, Briois, Sandrine, Cornic, Mathias, Elisee, Virginie, Denis, Jesuthasan, Djadi-Prat, Juliette, Jouany, Pauline, Junquera, Ramon, Henriques, Mickael, Kebir, Amina, Lehir, Isabelle, Meunier, Jeanne, Patin, Florence, Paquet, Val.rie, Tréhan, Anne, Vigna, Véronique, Sabatier, Brigitte, Bergerot, Damien, Jouve, Charléne, Knosp, Camille, Lenoir, Olivia, Mahtal, Nassim, Resmini, Léa, Lescure, Xavier, Ghosn, Jade, Bachelard, Antoine, Rachline, Anne, Isernia, Valentina, Bao-chau, Phung, Vallois, Dorothée, Sautereau, Aurelie, Neukrich, Catherine, Dossier, Antoine, Borie, Raphaël, Crestani, Bruno, Ducrocq, Gregory, Steg, Philippe Gabriel, Dieude, Philippe, Papo, Thomas, Marcault, Estelle, Chaudhry, Marhaba, Da Silveira, Charléne, Metois, Annabelle, Mahenni, Ismahan, Meziani, Meriam, Nilusmas, Cyndie, Le Gac, Sylvie, Ndiaye, Awa, Louni, Fran.oise, Chansombat, Malikhone, Julia, Zelie, Chalal, Solaya, Chalal, Lynda, Kramer, Laura, Le Grand, Jeniffer, Ouifiya, Kafif, Piquard, Valentine, Tubiana, Sarah, Nguyen, Yann, Honsel, Vasco, Weiss, Emmanuel, Codorniu, Anais, Zarrouk, Virginie, de Lastours, Victoire, Uzzan, Matthieu, Gamany, Naura, Claveirole, Agathe, Navid, Alexandre, Fouque, Tiffanie, Cohen, Yonathan, Lupo, Maya, Gilles, Constance, Rahli, Roza, Louis, Zeina, Boutboul, David, Galicier, Lionel, Amara, Yaël, Archer, Gabrielle, Benattia, Amira, Bergeron, Anne, Bondeelle, Louise, de Castro, Nathalie, Clément, Melissa, Darmon, Michaël, Denis, Blandine, Dupin, Clairelyne, Feredj, Elsa, Feyeux, Delphine, Joseph, Adrien, Lenglin, Etienne, Le Guen, Pierre, Liégeon, Geoffroy, Lorillon, Gwenaël, Mabrouki, Asma, Mariotte, Eric, Martin de Frémont, Grégoire, Mirouse, Adrien, Molina, Jean-Michel, Peffault de Latour, Régis, Oksenhendler, Eric, Saussereau, Julien, Tazi, Abdellatif, Tudesq, Jean-Jacques, Zafrani, Lara, Brindele, Isabelle, Bugnet, Emmanuelle, Lebras, Karine Celli, Chabert, Julien, Djaghout, Lamia, Fauvaux, Catherine, Jegu, Anne Lise, Kozakiewicz, Ewa, Meunier, Martine, Tremorin, Marie-Thérèse, Davoine, Claire, Madelaine, Isabelle, Caillat-Zucman, Sophie, Delaugerre, Constance, Morin, Florence, Sène, Damien, Burlacu, Ruxandra, Chousterman, Benjamin, Mégarbanne, Bruno, Richette, Pascal, Riveline, Jean-Pierre, Frazier, Aline, Vicaut, Eric, Berton, Laure, Hadjam, Tassadit, Vazquez-Ibarra, Miguel Alejandro, Jourdaine, Clément, Tran, Olivia, Jouis, Véronique, Jacob, Aude, Smati, Julie, Renaud, Stéphane, Pernin, Claire, Suarez, Lydia, Semerano, Luca, Abad, Sébastien, nainous, Ruben B., Bonnet, Nicolas, Comparon, Celine, Cohen, Yves, Cordel, Hugues, Dhote, Robin, Dournon, Nathalie, Duchemann, Boris, Ebstein, Nathan, Gille, Thomas, Giroux-Leprieur, Benedicte, Goupil de Bouille, Jeanne, Nunes, Hilario, Oziel, Johanna, Roulot, Dominique, Sese, Lucile, ClaireTantet, Uzunhan, Yurdagul, Bloch-Queyrat, Coralie, Levy, Vincent, Messani, Fadhila, Rahaoui, Mohammed, Petit, Myléne, Brahmi, Sabrina, Rathoin, Vanessa, Rigal, Marthe, Costedoat-Chalumeau, Nathalie, Luong, Liem Binh, Hamou, Zakaria Ait, Benghanem, Sarah, Blanche, Philippe, Carlier, Nicolas, Chaigne, Benjamin, Gauzit, Remy, Joumaa, Hassan, Jozwiak, Mathieu, Lachétre, Marie, Lafoeste, Hélène, Launay, Odie, Legendre, Paul, Marey, Jonathan, Morbieu, Caroline, Palmieri, Lola-Jade, Szwebel, Tali-Anne, Abdoul, Hendy, Bruneau, Alexandra, Beclin-Clabaux, Audrey, Larrieu, Charly, Montanari, Pierre, Dufour, Eric, Clarke, Ada, Le Bourlout, Catherine, Marin, Nathalie, Menage, Nathalie, Saleh-Mghir, Samira, Cisse, Mamadou Salif, Cheref, Kahina, Guerin, Corinne, Zerbit, Jérémie, Michel, Marc, Gallien, Sébastien, Crickx, Etienne, Le Vavasseur, Benjamin, Kempf, Emmanuelle, Jaffal, Karim, Vindrios, William, Oniszczuk, Julie, Guillaud, Constance, Lim, Pascal, Fois, Elena, Melica, Giovanna, Matignon, Marie, Jalabert, Maud, Lelièvre, Jean-Daniel, Schmitz, David, Bourhis, Marion, Belazouz, Sylia, Languille, Laetitia, Boucle, Caroline, Cita, Nelly, Didier, Agnés, Froura, Fahem, Ledudal, Katia, Sadaoui, Thiziri, Thiemele, Alaki, Le Febvre De Bailly, Delphine, Verlinde, Muriel Carvhalo, Mayaux, Julien, Cacoub, Patrice, Saadoun, David, Vautier, Mathieu, Bugaut, Héléne, Benveniste, Olivier, Allenbach, Yves, Leroux, Gaëlle, Rigolet, Aude, Guillaume-Jugnot, Perrine, Domont, Fanny, Desbois, Anne Claire, Comarmond, Chloé, Champtiaux, Nicolas, Toquet, Segolene, Ghembaza, Amine, Vieira, Matheus, Maalouf, Georgina, Boleto, Goncalo, Ferfar, Yasmina, Corvol, Jean-Christophe, Louapre, C.line, Sambin, Sara, Mariani, Louise-Laure, Karachi, Carine, Tubach, Florence, Estellat, Candice, Gimeno, Linda, Martin, Karine, Bah, Aicha, Keo, Vixra, Ouamri, Sabrine, Messaoudi, Yasmine, Yelles, Nessima, Faye, Pierre, Cavelot, Sebastien, Larcheveque, Cecile, Annonay, Laurence, Benhida, Jaouad, Zahrate-Ghoul, Aida, Hammal, Soumeya, Belilita, Ridha, Charbonnier, Fanny, Aguilar, Claire, Alby-Laurent, Fanny, Burger, Carole, Campos-Vega, Clara, Chavarot, Nathalie, Fournier, Benjamin, Rouzaud, Claire, Vimpére, Damien, Elie, Caroline, Bakouboula, Prissile, Choupeaux, Laure, Granville, Sophie, Issorat, Elodie, Broissand, Christine, Alyanakian, Marie-Alexandra, Geri, Guillaume, Derridj, Nawal, Sguiouar, Naima, Meddah, Hakim, Djadel, Mourad, Chambrin-Lauvray, Héléne, Duclos-vallée, Jean-Charles, Saliba, Faouzi, Sacleux, Sophie-Caroline, Kounis, Ilias, Tamazirt, Sonia, Rudant, Eric, Michot, Jean-Marie, Stoclin, Annabelle, Colomba, Emeline, Pommeret, Fanny, Willekens, Christophe, Da Silva, Rosa, Dejean, Valérie, Mekid, Yasmina, Ben-Mabrouk, Ines, Netzer, Florence, Pradon, Caroline, Drouard, Laurence, Camara-Clayette, Valérie, Morel, Alexandre, Garcia, Gilles, Mohebbi, Abolfazl, Berbour, Férial, Dehais, Mélanie, Pouliquen, Anne-Lise, Klasen, Alison, Soyez-Herkert, Loren, London, Jonathan, Keroumi, Younes, Guillot, Emmanuelle, Grailles, Guillaume, El amine, Younes, Defrancq, Fanny, Fodil, Hanane, Bouras, Chaouki, Dautel, Dominique, Gambier, Nicolas, Dieye, Thierno, Bienvenu, Boris, Lancon, Victor, Lecomte, Laurence, Beziriganyan, Kristina, Asselate, Belkacem, Allanic, Laure, Kiouris, Elena, Legros, Marie-Héléne, Lemagner, Christine, Martel, Pascal, Provitolo, Vincent, Ackermann, Félix, Le Marchand, Mathilde, Chan Hew Wai, Aurélie, Fremont, Dimitri, Coupez, Elisabeth, Adda, Mireille, Duée, Frédéric, Bernard, Lise, Gros, Antoine, Henry, Estelle, Courtin, Claire, Pattyn, Anne, Guinot, Pierre-Grégoire, Bardou, Marc, Maurer, Agnes, Jambon, Julie, Cransac, Amélie, Pernot, Corinne, Mourvillier, Bruno, Marquis, Eric, Benoit, Philippe, Roux, Damien, Gernez, Coralie, Yelnik, Cécile, Poissy, Julien, Nizard, Mandy, Denies, Fanette, Gros, Helene, Mourad, Jean-Jacques, Sacco, Emmanuelle, Renet, Sophie, Ader, F., Yazdanpanah, Y., Mentre, F., Peiffer-Smadja, N., Lescure, F.X., Poissy, J., Bouadma, L., Timsit, J.F., Lina, B., Morfin-Sherpa, F., Bouscambert, M., Gaymard, A., Peytavin, G., Abel, L., Guedj, J., Andrejak, C., Burdet, C., Laouenan, C., Belhadi, D., Dupont, A., Alfaiate, T., Basli, B., Chair, A., Laribi, S., Level, J., Schneider, M., Tellier, M.C., Dechanet, A., Costagliola, D., Terrier, B., Ohana, M., Couffin-Cadiergues, S., Esperou, H., Delmas, C., Saillard, J., Fougerou, C., Moinot, L., Wittkop, L., Cagnot, C., Le Mestre, S., Lebrasseur-Longuet, D., Petrov-Sanchez, V., Diallo, A., Mercier, N., Icard, V., Leveau, B., Tubiana, S., Hamze, B., Gelley, A., Noret, M., D’Ortenzio, E., Puechal, O., Semaille, C., Welte, T., Paiva, J.A., Halanova, M., Kieny, M.P., Balssa, E., Birkle, C., Gibowski, S., Landry, E., Le Goff, A., Moachon, L., Moins, C., Wadouachi, L., Paul, C., Levier, A., Bougon, D., Djossou, F., Epelboin, L., Dellamonica, J., Marquette, C.H., Robert, C., Gibot, S., Senneville, E., Jean-Michel, V., Zerbib, Y., Chirouze, C., Boyer, A., Cazanave, C., Gruson, D., Malvy, D., Andreu, P., Quenot, J.P., Terzi, N., Faure, K., Chabartier, C., Le Moing, V., Klouche, K., Ferry, T., F, Valour, Gaborit, B., Canet, E., Le Turnier, P., Boutoille, D., Bani-Sadr, F., Benezit, F., Revest, M., Cameli, C., Caro, A., Um Tegue, MJ Ngo, Le Tulzo, Y., Laviolle, B., Laine, F., Thiery, G., Meziani, F., Hansmann, Y., Oulehri, W., Tacquard, C., Vardon-Bounes, F., Riu-Poulenc, B., Murris-Espin, M., Bernard, L., Garot, D., Hinschberger, O., Martinot, M., Bruel, C., Pilmis, B., Bouchaud, O., Loubet, P., Roger, C., Monnet, X., Figueiredo, S., Godard, V., Mira, J.P., Lachatre, M., Kerneis, S., Aboab, J., Sayre, N., Crockett, F., Lebeaux, D., Buffet, A., Diehl, J.L., Fayol, A., Hulot, J.S., Livrozet, M., Dessap, A Mekontso, Ficko, C., Stefan, F., Le Pavec, J., Mayaux, J., Ait-Oufella, H., Molina, J.M., Pialoux, G., Fartoukh, M., Textoris, J., Brossard, M., Essat, A., Netzer, E., Riault, Y., Ghislain, M., Beniguel, L., Genin, M., Gouichiche, L., Betard, C., Belkhir, L., Altdorfer, A., Centro, V Fraipont, Braz, S., Ribeiro, JM Ferreira, Alburqueque, R Roncon, Berna, M., Alexandre, M., Lamprecht, B., Egle, A., Greil, R., Joannidis, M., Patterson, Thomas F., Ponce, Philip O., Taylor, Barbara S., Patterson, Jan E., Bowling, Jason E., Javeri, Heta, Kalil, Andre C., Larson, LuAnn, Hewlett, Angela, Mehta, Aneesh K., Rouphael, Nadine G., Saklawi, Youssef, Scanlon, Nicholas, Traenkner, Jessica J., Trible, Ronald P., Jr., Walter, Emmanuel B., Ivey, Noel, Holland, Thomas L., Ruiz-Palacios, Guillermo M., Ponce de León, Alfredo, Rajme, Sandra, Hsieh, Lanny, Amin, Alpesh N., Watanabe, Miki, Lee, Helen S., Kline, Susan, Billings, Joanne, Noren, Brooke, Kim, Hyun, Bold, Tyler D., Tapson, Victor, Grein, Jonathan, Sutterwala, Fayyaz, Iovine, Nicole, Beattie, Lars K., Wakeman, Rebecca Murray, Shaw, Matthew, Jain, Mamta K., Mocherla, Satish, Meisner, Jessica, Luque, Amneris, Sweeney, Daniel A., Benson, Constance A., Ali, Farhana, Atmar, Robert L., El Sahly, Hana M., Whitaker, Jennifer, Falsey, Ann R., Branche, Angela R., Rozario, Cheryl, Pineda, Justino Regalado, Martinez-Orozco, José Arturo, Lye, David Chien, Ong, Sean WX., Chia, Po Ying, Young, Barnaby E., Sandkovsky, Uriel, Berhe, Mezgebe, Haley, Clinton, Dishner, Emma, Cantos, Valeria D., Kelley, Colleen F., Rebolledo Esteinou, Paulina A., Kandiah, Sheetal, Doernberg, Sarah B., Crouch, Pierre-Cedric B., Jang, Hannah, Luetkemeyer, Anne F., Dwyer, Jay, Cohen, Stuart H., Thompson, George R., 3rd, Nguyen, Hien H., Finberg, Robert W., Wang, Jennifer P., Perez-Velazquez, Juan, Wessolossky, Mireya, Jackson, Patrick E.H., Bell, Taison D., West, Miranda J., Taiwo, Babafemi, Krueger, Karen, Perez, Johnny, Pearson, Triniece, Paules, Catharine I., Julian, Kathleen G., Ahmad, Danish, Hajduczok, Alexander G., Arguinchona, Henry, Arguinchona, Christa, Erdmann, Nathaniel, Goepfert, Paul, Ahuja, Neera, Frank, Maria G., Wyles, David, Young, Heather, Oh, Myoung-don, Park, Wan Beom, Kang, Chang Kyung, Marconi, Vincent, Moanna, Abeer, Cribbs, Sushma, Harrison, Telisha, Kim, Eu Suk, Jung, Jongtak, Song, Kyoung-Ho, Kim, Hong Bin, Tan, Seow Yen, Shafi, Humaira, Chien, Jaime, Fong, Raymond KC., Murray, Daniel D., Lundgren, Jens, Nielsen, Henrik, Jensen, Tomas, Zingman, Barry S., Grossberg, Robert, Riska, Paul F., Yang, Otto O., Ahn, Jenny, Arias, Rubi, Rapaka, Rekha R., Hauser, Naomi, Campbell, James D., Short, William R., Tebas, Pablo, Baron, Jillian T., McLellan, Susan L.F., Blanton, Lucas S., Seashore, Justin B., Creech, C. Buddy, Rice, Todd W., Walker, Shannon, Thomsen, Isaac P., Lopez de Castilla, Diego, Van Winkle, Jason W., Riedo, Francis X., Pada, Surinder Kaur, Wang, Alvin DY., Lin, Li, Harkins, Michelle, Mertz, Gregory, Sosa, Nestor, Ann Chai, Louis Yi, Tambyah, Paul Anantharajah, Tham, Sai Meng, Archuleta, Sophia, Yan, Gabriel, Lindholm, David A., Markelz, Ana Elizabeth, Mende, Katrin, Mularski, Richard, Hohmann, Elizabeth, Torres-Soto, Mariam, Jilg, Nikolaus, Maves, Ryan C., Utz, Gregory C., George, Sarah L., Hoft, Daniel F., Brien, James D., Paredes, Roger, Mateu, Lourdes, Loste, Cora, Kumar, Princy, Thornton, Sarah, Mohanraj, Sharmila, Hynes, Noreen A., Sauer, Lauren M., Colombo, Christopher J., Schofield, Christina, Colombo, Rhonda E., Chambers, Susan E., Novak, Richard M., Wendrow, Andrea, Gupta, Samir K., Lee, Tida, Lalani, Tahaniyat, Holodniy, Mark, Chary, Aarthi, Huprikar, Nikhil, Ganesan, Anuradha, Ohmagari, Norio, Mikami, Ayako, Price, D. Ashley, Duncan, Christopher J.A., Dierberg, Kerry, Neumann, Henry J., Taylor, Stephanie N., Lacour, Alisha, Masri, Najy, Swiatlo, Edwin, Widmer, Kyle, Neaton, James D., Bessesen, Mary, Stephens, David S., Burgess, Timothy H., Uyeki, Timothy M., Walker, Robert, Marks, G. Lynn, Osinusi, Anu, Cao, Huyen, Cardoso, Anabela, de Bono, Stephanie, Schlichting, Douglas E., Chung, Kevin K., Ferreira, Jennifer L., Green, Michelle, Makowski, Mat, Wierzbicki, Michael R., Conrad, Tom M., El-Khorazaty, Jill Ann, Hill, Heather, Bonnett, Tyler, Gettinger, Nikki, Engel, Theresa, Lewis, Teri, Wang, Jing, Beigel, John H., Tomashek, Kay M., Ghazaryan, Varduhi, Beresnev, Tatiana, Nayak, Seema, Dodd, Lori E., Dempsey, Walla, Nomicos, Effie, Lee, Marina, Pikaart-Tautges, Rhonda, Elsafy, Mohamed, Jurao, Robert, Koo, Hyung, Proschan, Michael, Yokum, Tammy, Arega, Janice, Florese, Ruth, Voell, Jocelyn D., Davey, Richard, Serrano, Ruth C., Wiley, Zanthia, Phadke, Varun K., Goepfert, Paul A., Gomez, Carlos A., Sofarelli, Theresa A., Certain, Laura, Imlay, Hannah N., Wolfe, Cameron R., Ko, Emily R., Engemann, John J., Felix, Nora Bautista, Wan, Claire R., Elmor, Sammy T., Bristow, Laurel R., Harkins, Michelle S., Iovine, Nicole M., Elie-Turenne, Marie-Carmelle, Tapson, Victor F., Choe, Pyoeng Gyun, Mularski, Richard A., Rhie, Kevin S., Hussein, Rezhan H., Ince, Dilek, Winokur, Patricia L., Takasaki, Jin, Saito, Sho, McConnell, Kimberly, Wyles, David L., Sarcone, Ellen, Grimes, Kevin A., Perez, Katherine, Janak, Charles, Whitaker, Jennifer A., Rebolledo, Paulina A., Gharbin, John, Lambert, Allison A., Zea, Diego F., Bainbridge, Emma, Hostler, David C., Hostler, Jordanna M., Shahan, Brian T., Ling, Evelyn, Go, Minjoung, Hubbard, Fleesie A., Chakrabarty, Melony, Laguio-Vila, Maryrose, Walsh, Edward E., Guirgis, Faheem, Marconi, Vincent C., Madar, Christian, Borgetti, Scott A., Levine, Corri, Nock, Joy, Candiotti, Keith, Rozman, Julia, Dangond, Fernando, Hyvert, Yann, Seitzinger, Andrea, Cross, Kaitlyn, Pettibone, Stephanie, Nayak, Seema U., Deye, Gregory A., Siempos, Ilias I., Belhadi, Drifa, Veiga, Viviane Cordeiro, Cavalcanti, Alexandre Biasi, Branch-Elliman, Westyn, Papoutsi, Eleni, Gkirgkiris, Konstantinos, Xixi, Nikoleta A., and Kotanidou, Anastasia

Published
2024
Full Text
View/download PDF

18. Efficacy of interferon beta-1a plus remdesivir compared with remdesivir alone in hospitalised adults with COVID-19: a double-bind, randomised, placebo-controlled, phase 3 trial

Author

Kalil, Andre C, Mehta, Aneesh K, Patterson, Thomas F, Erdmann, Nathaniel, Gomez, Carlos A, Jain, Mamta K, Wolfe, Cameron R, Ruiz-Palacios, Guillermo M, Kline, Susan, Pineda, Justino Regalado, Luetkemeyer, Anne F, Harkins, Michelle S, Jackson, Patrick EH, Iovine, Nicole M, Tapson, Victor F, Oh, Myoung-don, Whitaker, Jennifer A, Mularski, Richard A, Paules, Catharine I, Ince, Dilek, Takasaki, Jin, Sweeney, Daniel A, Sandkovsky, Uriel, Wyles, David L, Hohmann, Elizabeth, Grimes, Kevin A, Grossberg, Robert, Laguio-Vila, Maryrose, Lambert, Allison A, de Castilla, Diego Lopez, Kim, EuSuk, Larson, LuAnn, Wan, Claire R, Traenkner, Jessica J, Ponce, Philip O, Patterson, Jan E, Goepfert, Paul A, Sofarelli, Theresa A, Mocherla, Satish, Ko, Emily R, de Leon, Alfredo Ponce, Doernberg, Sarah B, Atmar, Robert L, Maves, Ryan C, Dangond, Fernando, Ferreira, Jennifer, Green, Michelle, Makowski, Mat, Bonnett, Tyler, Beresnev, Tatiana, Ghazaryan, Varduhi, Dempsey, Walla, Nayak, Seema U, Dodd, Lori, Tomashek, Kay M, Beigel, John H, members, ACTT-3 study group, Hewlett, Angela, Taylor, Barbara S, Bowling, Jason E, Serrano, Ruth C, Rouphael, Nadine G, Wiley, Zanthia, Phadke, Varun K, Certain, Laura, Imlay, Hannah N, Engemann, John J, Walter, Emmanuel B, Meisner, Jessica, Rajme, Sandra, Billings, Joanne, Kim, Hyun, Martinez-Orozco, Jose A, Felix, Nora Bautista, Elmor, Sammy T, Bristow, Laurel R, Mertz, Gregory, Sosa, Nestor, Bell, Taison D, West, Miranda J, Elie-Turenne, Marie-Carmelle, Grein, Jonathan, Sutterwala, Fayyaz, Choe, Pyoeng Gyun, Kang, Chang Kyung, Sahly, Hana M El, Rhie, Kevin S, Hussein, Rezhan H, Winokur, Patricia L, Mikami, Ayako, Saito, Sho, Benson, Constance A, McConnell, Kimberly, Berhe, Mezgebe, Dishner, Emma, Frank, Maria G, Sarcone, Ellen, Crouch, Pierre-Cedric B, Jang, Hannah, and Jilg, Nikolaus

Subjects
Rehabilitation, Clinical Trials and Supportive Activities, Lung, Clinical Research, Infectious Diseases, 6.1 Pharmaceuticals, Evaluation of treatments and therapeutic interventions, Infection, Adenosine Monophosphate, Adult, Aged, Alanine, Antiviral Agents, Double-Blind Method, Female, Humans, Interferon beta-1a, Japan, Male, Mexico, Middle Aged, Oxygen, Oxygen Saturation, Republic of Korea, SARS-CoV-2, Singapore, Treatment Outcome, United States, COVID-19 Drug Treatment, ACTT-3 study group members, Clinical Sciences, Public Health and Health Services, Other Medical and Health Sciences
Abstract

BackgroundFunctional impairment of interferon, a natural antiviral component of the immune system, is associated with the pathogenesis and severity of COVID-19. We aimed to compare the efficacy of interferon beta-1a in combination with remdesivir compared with remdesivir alone in hospitalised patients with COVID-19.MethodsWe did a double-blind, randomised, placebo-controlled trial at 63 hospitals across five countries (Japan, Mexico, Singapore, South Korea, and the USA). Eligible patients were hospitalised adults (aged ≥18 years) with SARS-CoV-2 infection, as confirmed by a positive RT-PCR test, and who met one of the following criteria suggestive of lower respiratory tract infection: the presence of radiographic infiltrates on imaging, a peripheral oxygen saturation on room air of 94% or less, or requiring supplemental oxygen. Patients were excluded if they had either an alanine aminotransferase or an aspartate aminotransferase concentration more than five times the upper limit of normal; had impaired renal function; were allergic to the study product; were pregnant or breast feeding; were already on mechanical ventilation; or were anticipating discharge from the hospital or transfer to another hospital within 72 h of enrolment. Patients were randomly assigned (1:1) to receive intravenous remdesivir as a 200 mg loading dose on day 1 followed by a 100 mg maintenance dose administered daily for up to 9 days and up to four doses of either 44 μg interferon beta-1a (interferon beta-1a group plus remdesivir group) or placebo (placebo plus remdesivir group) administered subcutaneously every other day. Randomisation was stratified by study site and disease severity at enrolment. Patients, investigators, and site staff were masked to interferon beta-1a and placebo treatment; remdesivir treatment was given to all patients without masking. The primary outcome was time to recovery, defined as the first day that a patient attained a category 1, 2, or 3 score on the eight-category ordinal scale within 28 days, assessed in the modified intention-to-treat population, defined as all randomised patients who were classified according to actual clinical severity. Safety was assessed in the as-treated population, defined as all patients who received at least one dose of the assigned treatment. This trial is registered with ClinicalTrials.gov, NCT04492475.FindingsBetween Aug 5, 2020, and Nov 11, 2020, 969 patients were enrolled and randomly assigned to the interferon beta-1a plus remdesivir group (n=487) or to the placebo plus remdesivir group (n=482). The mean duration of symptoms before enrolment was 8·7 days (SD 4·4) in the interferon beta-1a plus remdesivir group and 8·5 days (SD 4·3) days in the placebo plus remdesivir group. Patients in both groups had a time to recovery of 5 days (95% CI not estimable) (rate ratio of interferon beta-1a plus remdesivir group vs placebo plus remdesivir 0·99 [95% CI 0·87-1·13]; p=0·88). The Kaplan-Meier estimate of mortality at 28 days was 5% (95% CI 3-7%) in the interferon beta-1a plus remdesivir group and 3% (2-6%) in the placebo plus remdesivir group (hazard ratio 1·33 [95% CI 0·69-2·55]; p=0·39). Patients who did not require high-flow oxygen at baseline were more likely to have at least one related adverse event in the interferon beta-1a plus remdesivir group (33 [7%] of 442 patients) than in the placebo plus remdesivir group (15 [3%] of 435). In patients who required high-flow oxygen at baseline, 24 (69%) of 35 had an adverse event and 21 (60%) had a serious adverse event in the interferon beta-1a plus remdesivir group compared with 13 (39%) of 33 who had an adverse event and eight (24%) who had a serious adverse event in the placebo plus remdesivir group.InterpretationInterferon beta-1a plus remdesivir was not superior to remdesivir alone in hospitalised patients with COVID-19 pneumonia. Patients who required high-flow oxygen at baseline had worse outcomes after treatment with interferon beta-1a compared with those given placebo.FundingThe National Institute of Allergy and Infectious Diseases (USA).

Published
2021

19. GIST: Distributed Training for Large-Scale Graph Convolutional Networks

Author

Wolfe, Cameron R., Yang, Jingkang, Chowdhury, Arindam, Dun, Chen, Bayer, Artun, Segarra, Santiago, and Kyrillidis, Anastasios

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Distributed, Parallel, and Cluster Computing, Mathematics - Optimization and Control, I.2.4
Abstract

The graph convolutional network (GCN) is a go-to solution for machine learning on graphs, but its training is notoriously difficult to scale both in terms of graph size and the number of model parameters. Although some work has explored training on large-scale graphs (e.g., GraphSAGE, ClusterGCN, etc.), we pioneer efficient training of large-scale GCN models (i.e., ultra-wide, overparameterized models) with the proposal of a novel, distributed training framework. Our proposed training methodology, called GIST, disjointly partitions the parameters of a GCN model into several, smaller sub-GCNs that are trained independently and in parallel. In addition to being compatible with all GCN architectures and existing sampling techniques for efficient GCN training, GIST i) improves model performance, ii) scales to training on arbitrarily large graphs, iii) decreases wall-clock training time, and iv) enables the training of markedly overparameterized GCN models. Remarkably, with GIST, we train an astonishgly-wide 32,768-dimensional GraphSAGE model, which exceeds the capacity of a single GPU by a factor of 8x, to SOTA performance on the Amazon2M dataset., Comment: 28 pages, 5 figures, pre-print under review

Published
2021

20. Remdesivir for the prevention of invasive mechanical ventilation or death in COVID-19 - A post-hoc analysis of the Adaptive COVID-19 Treatment Trial-1 Cohort Data.

Author

Paules, Catharine I, Gallagher, Shannon K, Rapaka, Rekha R, Davey, Richard T, Doernberg, Sarah B, Grossberg, Robert, Hynes, Noreen A, Ponce, Philip, Short, William R, Voell, Jocelyn, Wang, Jing, Yang, Otto O, Wolfe, Cameron R, Lye, David C, Dodd, Lori E, and Benson, Constance A

Subjects
ACTT-1, COVID-19, Remdesivir, Lung, Prevention, Rare Diseases, Microbiology, Biological Sciences, Medical and Health Sciences
Abstract

This post-hoc analysis of The Adaptive COVID-19 Treatment Trial-1 shows a treatment effect of remdesivir on progression to invasive mechanical ventilation (IMV) or death. Additionally, we create a risk profile that better predicts progression than baseline oxygen requirement alone. The highest risk group derives the greatest treatment effect from RDV.

Published
2021

21. Baricitinib plus Remdesivir for Hospitalized Adults with Covid-19

Author

Kalil, Andre C, Patterson, Thomas F, Mehta, Aneesh K, Tomashek, Kay M, Wolfe, Cameron R, Ghazaryan, Varduhi, Marconi, Vincent C, Ruiz-Palacios, Guillermo M, Hsieh, Lanny, Kline, Susan, Tapson, Victor, Iovine, Nicole M, Jain, Mamta K, Sweeney, Daniel A, El Sahly, Hana M, Branche, Angela R, Regalado Pineda, Justino, Lye, David C, Sandkovsky, Uriel, Luetkemeyer, Anne F, Cohen, Stuart H, Finberg, Robert W, Jackson, Patrick EH, Taiwo, Babafemi, Paules, Catharine I, Arguinchona, Henry, Erdmann, Nathaniel, Ahuja, Neera, Frank, Maria, Oh, Myoung-Don, Kim, Eu-Suk, Tan, Seow Y, Mularski, Richard A, Nielsen, Henrik, Ponce, Philip O, Taylor, Barbara S, Larson, LuAnn, Rouphael, Nadine G, Saklawi, Youssef, Cantos, Valeria D, Ko, Emily R, Engemann, John J, Amin, Alpesh N, Watanabe, Miki, Billings, Joanne, Elie, Marie-Carmelle, Davey, Richard T, Burgess, Timothy H, Ferreira, Jennifer, Green, Michelle, Makowski, Mat, Cardoso, Anabela, de Bono, Stephanie, Bonnett, Tyler, Proschan, Michael, Deye, Gregory A, Dempsey, Walla, Nayak, Seema U, Dodd, Lori E, and Beigel, John H

Subjects
Rehabilitation, Clinical Trials and Supportive Activities, Lung, Clinical Research, 6.1 Pharmaceuticals, Evaluation of treatments and therapeutic interventions, Good Health and Well Being, Adenosine Monophosphate, Adult, Aged, Alanine, Antiviral Agents, Azetidines, COVID-19, Double-Blind Method, Drug Therapy, Combination, Female, Hospital Mortality, Hospitalization, Humans, Janus Kinase Inhibitors, Male, Middle Aged, Oxygen Inhalation Therapy, Purines, Pyrazoles, Respiration, Artificial, Sulfonamides, Treatment Outcome, COVID-19 Drug Treatment, ACTT-2 Study Group Members, Medical and Health Sciences, General & Internal Medicine
Abstract

BackgroundSevere coronavirus disease 2019 (Covid-19) is associated with dysregulated inflammation. The effects of combination treatment with baricitinib, a Janus kinase inhibitor, plus remdesivir are not known.MethodsWe conducted a double-blind, randomized, placebo-controlled trial evaluating baricitinib plus remdesivir in hospitalized adults with Covid-19. All the patients received remdesivir (≤10 days) and either baricitinib (≤14 days) or placebo (control). The primary outcome was the time to recovery. The key secondary outcome was clinical status at day 15.ResultsA total of 1033 patients underwent randomization (with 515 assigned to combination treatment and 518 to control). Patients receiving baricitinib had a median time to recovery of 7 days (95% confidence interval [CI], 6 to 8), as compared with 8 days (95% CI, 7 to 9) with control (rate ratio for recovery, 1.16; 95% CI, 1.01 to 1.32; P = 0.03), and a 30% higher odds of improvement in clinical status at day 15 (odds ratio, 1.3; 95% CI, 1.0 to 1.6). Patients receiving high-flow oxygen or noninvasive ventilation at enrollment had a time to recovery of 10 days with combination treatment and 18 days with control (rate ratio for recovery, 1.51; 95% CI, 1.10 to 2.08). The 28-day mortality was 5.1% in the combination group and 7.8% in the control group (hazard ratio for death, 0.65; 95% CI, 0.39 to 1.09). Serious adverse events were less frequent in the combination group than in the control group (16.0% vs. 21.0%; difference, -5.0 percentage points; 95% CI, -9.8 to -0.3; P = 0.03), as were new infections (5.9% vs. 11.2%; difference, -5.3 percentage points; 95% CI, -8.7 to -1.9; P = 0.003).ConclusionsBaricitinib plus remdesivir was superior to remdesivir alone in reducing recovery time and accelerating improvement in clinical status among patients with Covid-19, notably among those receiving high-flow oxygen or noninvasive ventilation. The combination was associated with fewer serious adverse events. (Funded by the National Institute of Allergy and Infectious Diseases; ClinicalTrials.gov number, NCT04401579.).

Published
2021

22. E-Stitchup: Data Augmentation for Pre-Trained Embeddings

Author

Wolfe, Cameron R. and Lundgaard, Keld T.

Subjects
Computer Science - Machine Learning, Statistics - Machine Learning
Abstract

In this work, we propose data augmentation methods for embeddings from pre-trained deep learning models that take a weighted combination of a pair of input embeddings, as inspired by Mixup, and combine such augmentation with extra label softening. These methods are shown to significantly increase classification accuracy, reduce training time, and improve confidence calibration of a downstream model that is trained with them. As a result of such improved confidence calibration, the model output can be more intuitively interpreted and used to accurately identify out-of-distribution data by applying an appropriate confidence threshold to model predictions. The identified out-of-distribution data can then be prioritized for labeling, thus focusing labeling effort on data that is more likely to boost model performance. These findings, we believe, lay a solid foundation for improving the classification performance and calibration of models that use pre-trained embeddings as input and provide several benefits that prove extremely useful in a production-level deep learning system., Comment: 11 pages, 7 figures

Published
2019

23. Distributed Learning of Deep Neural Networks using Independent Subnet Training

Author

Yuan, Binhang, Wolfe, Cameron R., Dun, Chen, Tang, Yuxin, Kyrillidis, Anastasios, and Jermaine, Christopher M.

Subjects
Computer Science - Machine Learning, Statistics - Machine Learning
Abstract

Distributed machine learning (ML) can bring more computational resources to bear than single-machine learning, thus enabling reductions in training time. Distributed learning partitions models and data over many machines, allowing model and dataset sizes beyond the available compute power and memory of a single machine. In practice though, distributed ML is challenging when distribution is mandatory, rather than chosen by the practitioner. In such scenarios, data could unavoidably be separated among workers due to limited memory capacity per worker or even because of data privacy issues. There, existing distributed methods will utterly fail due to dominant transfer costs across workers, or do not even apply. We propose a new approach to distributed fully connected neural network learning, called independent subnet training (IST), to handle these cases. In IST, the original network is decomposed into a set of narrow subnetworks with the same depth. These subnetworks are then trained locally before parameters are exchanged to produce new subnets and the training cycle repeats. Such a naturally "model parallel" approach limits memory usage by storing only a portion of network parameters on each device. Additionally, no requirements exist for sharing data between workers (i.e., subnet training is local and independent) and communication volume and frequency are reduced by decomposing the original network into independent subnets. These properties of IST can cope with issues due to distributed data, slow interconnects, or limited device memory, making IST a suitable approach for cases of mandatory distribution. We show experimentally that IST results in training times that are much lower than common distributed learning approaches.

Published
2019

24. Living kidney donors with HIV: experience and outcomes from a case series by the HOPE in Action Consortium

Author

Agrawal, Neerja, Pereira, Marcus, Ranganna, Karthik, Wolfe, Cameron, Friedman-Moraco, Rachel, Kitchens, William, Adebiyi, Oluwafisayo, Kubal, Chandrashekhar, Cameron, Andrew, Desai, Niraj, Durand, Christine, Ottmann, Shane, Elias, Nahel, Gilbert, Alexander, Smith, Coleman, Castillo-Lugo, Jose A., Florman, Sander, Segev, Dorry L., Massie, Allan, Mehta, Sapna, Stosor, Valentina, Hand, Jonathan, Blumberg, Emily, Santos, Carlos A.Q., Goldberg, Ryan, Mehta, Shikha, Cannon, Robert, Giorgakis, Emmanouil, Schaenman, Joanna, Aslam, Saima, Stock, Peter, Price, Jennifer, Apewokin, Senu, Benamu, Esther, Spaggiari, Mario, Baddley, John, Morris, Michele I., Simkins, Jacques, Pruett, Timothy, Haidar, Ghady, Wojciechowski, David, Agarwal, Avinash, Balaraman, Vasanthi, Gupta, Gaurav, Chapman, Will, Muthukumar, Thangamani, Small, Catherine B., Malinis, Maricar, Durand, Christine M., Martinez, Nina, Neumann, Karl, Benedict, Reed C., Baker, Arthur W., Wolfe, Cameron R., Shetty, Aneesha, Dietch, Zachary C., Goudy, Leah, Callegari, Michelle A., Massie, Allan B., Brown, Diane, Cochran, Willa, Muzaale, Abimereki, Fine, Derek, Tobian, Aaron A.R., Winkler, Cheryl A., and Al Ammary, Fawaz

Published
2023
Full Text
View/download PDF

25. Functional Generative Design of Mechanisms with Recurrent Neural Networks and Novelty Search

Author

Wolfe, Cameron R., Tutum, Cem C., and Miikkulainen, Risto

Subjects
Computer Science - Neural and Evolutionary Computing
Abstract

Consumer-grade 3D printers have made it easier to fabricate aesthetic objects and static assemblies, opening the door to automated design of such objects. However, while static designs are easily produced with 3D printing, functional designs with moving parts are more difficult to generate: The search space is too high-dimensional, the resolution of the 3D-printed parts is not adequate, and it is difficult to predict the physical behavior of imperfect 3D-printed mechanisms. An example challenge is to produce a diverse set of reliable and effective gear mechanisms that could be used after production without extensive post-processing. To meet this challenge, an indirect encoding based on a Recurrent Neural Network (RNN) is created and evolved using novelty search. The elite solutions of each generation are 3D printed to evaluate their functional performance on a physical test platform. The system is able to discover sequential design rules that are difficult to discover with other methods. Compared to direct encoding evolved with Genetic Algorithms (GAs), its designs are geometrically more diverse and functionally more effective. It therefore forms a promising foundation for the generative design of 3D-printed, functional mechanisms., Comment: 7 pages, GECCO 2019

Published
2019

48. Baricitinib versus dexamethasone for adults hospitalised with COVID-19 (ACTT-4): a randomised, double-blind, double placebo-controlled trial

Author

Wahid, Lana, Walter, Emmanuel B., Belur, Akhila G., Dreyer, Grace, Patterson, Jan E., Bowling, Jason E., Dixon, Danielle O., Hewlett, Angela, Odrobina, Robert, Pupaibool, Jakrapun, Mocherla, Satish, Lazarte, Suzana, Cayabyab, Meilani, Hussein, Rezhan H., Golamari, Reshma R., Krill, Kaleigh L., Rajme, Sandra, Riska, Paul F., Zingman, Barry S., Mertz, Gregory, Sosa, Nestor, Goepfert, Paul A., Berhe, Mezgebe, Dishner, Emma, Fayed, Mohamed, Hubel, Kinsley, Martinez-Orozco, José Arturo, Bautista Felix, Nora, Elmor, Sammy T., Bechnak, Amer Ryan, Saklawi, Youssef, Van Winkle, Jason W., Zea, Diego F., Laguio-Vila, Maryrose, Walsh, Edward E., Falsey, Ann R., Carvajal, Karen, Hyzy, Robert C., Hanna, Sinan, Olbrich, Norman, Traenkner, Jessica J., Kraft, Colleen S., Tebas, Pablo, Baron, Jillian T, Levine, Corri, Nock, Joy, Billings, Joanne, Kim, Hyun, Elie-Turenne, Marie-Carmelle, Whitaker, Jennifer A., Luetkemeyer, Anne F., Dwyer, Jay, Bainbridge, Emma, Gyun Choe, Pyoeng, Kyung Kang, Chang, Jilg, Nikolaus, Cantos, Valeria D, Bhamidipati, Divya R., Nithin Gopalsamy, Srinivasa, Chary, Aarthi, Jung, Jongtak, Song, Kyoung-Ho, Kim, Hong Bin, Benson, Constance A., McConnell, Kimberly, Wang, Jennifer P., Wessolossky, Mireya, Perez, Katherine, Eubank, Taryn A, Berjohn, Catherine, Utz, Gregory C., Jackson, Patrick E.H., Bell, Taison D., Haughey, Heather M., Moanna, Abeer, Cribbs, Sushma, Harrison, Telisha, Colombo, Christopher J., Schofield, Christina, Colombo, Rhonda E., Tapson, Victor F., Grein, Jonathan, Sutterwala, Fayyaz, Ince, Dilek, Winokur, Patricia L., Fung, Monica, Jang, Hannah, Wyles, David, Frank, Maria G., Sarcone, Ellen, Neumann, Henry, Viswanathan, Anand, Hochman, Sarah, Mulligan, Mark, Eckhardt, Benjamin, Carmody, Ellie, Ahuja, Neera, Nadeau, Kari, Svec, David, Macaraeg, Jeffrey C., Morrow, Lee, Quimby, Dave, Bessesen, Mary, Nicholson, Lindsay, Adams, Jill, Kumar, Princy, Lambert, Allison A., Arguinchona, Henry, Alicic, Radica Z., Saito, Sho, Ohmagari, Norio, Mikami, Ayako, Chien Lye, David, Hong Lee, Tau, Ying Chia, Po, Hsieh, Lanny, Amin, Alpesh N., Watanabe, Miki, Candiotti, Keith A., Castro, Jose G., Antor, Maria A., Lee, Tida, Lalani, Tahaniyat, Novak, Richard M., Wendrow, Andrea, Borgetti, Scott A., George, Sarah L., Hoft, Daniel F., Brien, James D., Cohen, Stuart H., Thompson, George R., 3rd, Chakrabarty, Melony, Guirgis, Faheem, Davey, Richard T., Voell, Jocelyn, Strich, Jeffrey R., Lindholm, David A., Mende, Katrin, Wellington, Trevor R., Rapaka, Rekha R., Husson, Jennifer S., Levine, Andrea R., Yen Tan, Seow, Shafi, Humaira, Chien, Jaime M F, Hostler, David C., Hostler, Jordanna M., Shahan, Brian T., Adams, David H., Osinusi, Anu, Cao, Huyen, Burgess, Timothy H., Rozman, Julia, Chung, Kevin K., Nieuwoudt, Christina, El-Khorazaty, Jill A., Hill, Heather, Pettibone, Stephanie, Gettinger, Nikki, Engel, Theresa, Lewis, Teri, Wang, Jing, Deye, Gregory A., Nomicos, Effie, Pikaart-Tautges, Rhonda, Elsafy, Mohamed, Jurao, Robert, Koo, Hyung, Proschan, Michael, Yokum, Tammy, Arega, Janice, Florese, Ruth, Wolfe, Cameron R, Tomashek, Kay M, Patterson, Thomas F, Gomez, Carlos A, Marconi, Vincent C, Jain, Mamta K, Yang, Otto O, Paules, Catharine I, Palacios, Guillermo M Ruiz, Grossberg, Robert, Harkins, Michelle S, Mularski, Richard A, Erdmann, Nathaniel, Sandkovsky, Uriel, Almasri, Eyad, Pineda, Justino Regalado, Dretler, Alexandra W, de Castilla, Diego Lopez, Branche, Angela R, Park, Pauline K, Mehta, Aneesh K, Short, William R, McLellan, Susan L F, Kline, Susan, Iovine, Nicole M, El Sahly, Hana M, Doernberg, Sarah B, Oh, Myoung-don, Huprikar, Nikhil, Hohmann, Elizabeth, Kelley, Colleen F, Holodniy, Mark, Kim, Eu Suk, Sweeney, Daniel A, Finberg, Robert W, Grimes, Kevin A, Maves, Ryan C, Ko, Emily R, Engemann, John J, Taylor, Barbara S, Ponce, Philip O, Larson, LuAnn, Melendez, Dante Paolo, Seibert, Allan M, Rouphael, Nadine G, Strebe, Joslyn, Clark, Jesse L, Julian, Kathleen G, de Leon, Alfredo Ponce, Cardoso, Anabela, de Bono, Stephanie, Atmar, Robert L, Ganesan, Anuradha, Ferreira, Jennifer L, Green, Michelle, Makowski, Mat, Bonnett, Tyler, Beresnev, Tatiana, Ghazaryan, Varduhi, Dempsey, Walla, Nayak, Seema U, Dodd, Lori E, Beigel, John H, and Kalil, Andre C

Published
2022
Full Text
View/download PDF

50. Efficacy of interferon beta-1a plus remdesivir compared with remdesivir alone in hospitalised adults with COVID-19: a double-blind, randomised, placebo-controlled, phase 3 trial

Author

Hewlett, Angela, Taylor, Barbara S, Bowling, Jason E, Serrano, Ruth C, Rouphael, Nadine G, Wiley, Zanthia, Phadke, Varun K, Certain, Laura, Imlay, Hannah N, Engemann, John J, Walter, Emmanuel B, Meisner, Jessica, Rajme, Sandra, Billings, Joanne, Kim, Hyun, Martinez-Orozco, Jose A, Bautista Felix, Nora, Elmor, Sammy T, Bristow, Laurel R, Mertz, Gregory, Sosa, Nestor, Bell, Taison D, West, Miranda J, Elie-Turenne, Marie-Carmelle, Grein, Jonathan, Sutterwala, Fayyaz, Gyun Choe, Pyoeng, Kyung Kang, Chang, El Sahly, Hana M, Rhie, Kevin S, Hussein, Rezhan H, Winokur, Patricia L, Mikami, Ayako, Saito, Sho, Benson, Constance A, McConnell, Kimberly, Berhe, Mezgebe, Dishner, Emma, Frank, Maria G, Sarcone, Ellen, Crouch, Pierre-Cedric B, Jang, Hannah, Jilg, Nikolaus, Perez, Katherine, Janak, Charles, Cantos, Valeria D, Rebolledo, Paulina A, Gharbin, John, Zingman, Barry S, Riska, Paul F, Falsey, Ann R, Walsh, Edward E, Branche, Angela R, Arguinchona, Henry, Arguinchona, Christa, Van Winkle, Jason W, Zea, Diego F, Jung, Jongtak, Song, Kyoung-Ho, Kim, Hong Bin, Dwyer, Jay, Bainbridge, Emma, Hostler, David C, Hostler, Jordanna M, Shahan, Brian T, Hsieh, Lanny, Amin, Alpesh N, Watanabe, Miki, Short, William R, Tebas, Pablo, Baron, Jillian T, Ahuja, Neera, Ling, Evelyn, Go, Minjoung, Yang, Otto O, Ahn, Jenny, Arias, Rubi, Rapaka, Rekha R, Hubbard, Fleesie A, Campbell, James D, Cohen, Stuart H, Thompson, George R, 3rd, Chakrabarty, Melony, Taylor, Stephanie N, Masri, Najy, Lacour, Alisha, Lee, Tida, Lalani, Tahaniyat, Lindholm, David A, Markelz, Ana Elizabeth, Mende, Katrin, Colombo, Christopher J, Schofield, Christina, Colombo, Rhonda E, Guirgis, Faheem, Holodniy, Mark, Chary, Aarthi, Bessesen, Mary, Hynes, Noreen A, Sauer, Lauren M, Marconi, Vincent C, Moanna, Abeer, Harrison, Telisha, Lye, David C, Ong, Sean W X, Ying Chia, Po, Huprikar, Nikhil, Ganesan, Anuradha, Madar, Christian, Novak, Richard M, Wendrow, Andrea, Borgetti, Scott A, George, Sarah L, Hoft, Daniel F, Brien, James D, McLellan, Susan L F, Levine, Corri, Nock, Joy, Yen Tan, Seow, Shafi, Humaira, Chien, Jaime M F, Candiotti, Keith, Finberg, Robert W, Wang, Jennifer P, Wessolossky, Mireya, Utz, Gregory C, Chambers, Susan E, Stephens, David S, Burgess, Timothy H, Rozman, Julia, Hyvert, Yann, Seitzinger, Andrea, Osinusi, Anu, Cao, Huyen, Chung, Kevin K, Conrad, Tom M, Cross, Kaitlyn, El-Khorazaty, Jill A, Hill, Heather, Pettibone, Stephanie, Wierzbicki, Michael R, Gettinger, Nikki, Engel, Theresa, Lewis, Teri, Wang, Jing, Deye, Gregory A, Nomicos, Effie, Pikaart-Tautges, Rhonda, Elsafy, Mohamed, Jurao, Robert, Koo, Hyung, Proschan, Michael, Davey, Richard, Yokum, Tammy, Arega, Janice, Florese, Ruth, Kalil, Andre C, Mehta, Aneesh K, Patterson, Thomas F, Erdmann, Nathaniel, Gomez, Carlos A, Jain, Mamta K, Wolfe, Cameron R, Ruiz-Palacios, Guillermo M, Kline, Susan, Regalado Pineda, Justino, Luetkemeyer, Anne F, Harkins, Michelle S, Jackson, Patrick E H, Iovine, Nicole M, Tapson, Victor F, Oh, Myoung-don, Whitaker, Jennifer A, Mularski, Richard A, Paules, Catharine I, Ince, Dilek, Takasaki, Jin, Sweeney, Daniel A, Sandkovsky, Uriel, Wyles, David L, Hohmann, Elizabeth, Grimes, Kevin A, Grossberg, Robert, Laguio-Vila, Maryrose, Lambert, Allison A, Lopez de Castilla, Diego, Kim, EuSuk, Larson, LuAnn, Wan, Claire R, Traenkner, Jessica J, Ponce, Philip O, Patterson, Jan E, Goepfert, Paul A, Sofarelli, Theresa A, Mocherla, Satish, Ko, Emily R, Ponce de Leon, Alfredo, Doernberg, Sarah B, Atmar, Robert L, Maves, Ryan C, Dangond, Fernando, Ferreira, Jennifer, Green, Michelle, Makowski, Mat, Bonnett, Tyler, Beresnev, Tatiana, Ghazaryan, Varduhi, Dempsey, Walla, Nayak, Seema U, Dodd, Lori, Tomashek, Kay M, and Beigel, John H

Published
2021
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results