Eduardo Ponce Mojica
@utk.edu
Graduate Research Assistant in Electrical Engineering and Computer Science Department
The University of Tennessee, Knoxville
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Brittany Stephenson, Cristina Lanzas, Suzanne Lenhart, Eduardo Ponce, Jason Bintz, Erik R. Dubberke, and Judy Day
Springer Science and Business Media LLC
Abstract Background Clostridioides difficile infection (CDI) is one of the most common healthcare infections. Common strategies aiming at controlling CDI include antibiotic stewardship, environmental decontamination, and improved hand hygiene and contact precautions. Mathematical models provide a framework to evaluate control strategies. Our objective is to evaluate the effectiveness of control strategies in decreasing C. difficile colonization and infection using an agent-based model in an acute healthcare setting. Methods We developed an agent-based model that simulates the transmission of C. difficile in medical wards. This model explicitly incorporates healthcare workers (HCWs) as vectors of transmission, tracks individual patient antibiotic histories, incorporates varying risk levels of antibiotics with respect to CDI susceptibility, and tracks contamination levels of ward rooms by C. difficile. Interventions include two forms of antimicrobial stewardship, increased environmental decontamination through room cleaning, improved HCW compliance, and a preliminary assessment of vaccination. Results Increased HCW compliance with CDI patients was ranked as the most effective intervention in decreasing colonizations, with reductions up to 56%. Antibiotic stewardship practices were highly ranked after contact precaution compliance. Vaccination and reduction of high-risk antibiotics were the most effective intervention in decreasing CDI. Vaccination reduced CDI cases to up to 90%, and the reduction of high-risk antibiotics decreased CDI cases up to 23%. Conclusions Overall, interventions that decrease patient susceptibility to colonization by C. difficile, such as antibiotic stewardship, were the most effective interventions in reducing both colonizations and CDI cases.
Tapajit Dey, Sara Mousavi, Eduardo Ponce, Tanner Fry, Bogdan Vasilescu, Anna Filippova, and Audris Mockus
ACM
Background: Some developer activity traditionally performed manually, such as making code commits, opening, managing, or closing issues is increasingly subject to automation in many OSS projects. Specifically, such activity is often performed by tools that react to events or run at specific times. We refer to such automation tools as bots and, in many software mining scenarios related to developer productivity or code quality, it is desirable to identify bots in order to separate their actions from actions of individuals. Aim: Find an automated way of identifying bots and code committed by these bots, and to characterize the types of bots based on their activity patterns. Method and Result: We propose BIMAN, a systematic approach to detect bots using author names, commit messages, files modified by the commit, and projects associated with the commits. For our test data, the value for AUC-ROC was 0.9. We also characterized these bots based on the time patterns of their code commits and the types of files modified, and found that they primarily work with documentation files and web pages, and these files are most prevalent in HTML and JavaScript ecosystems. We have compiled a shareable dataset containing detailed information about 461 bots we found (all of which have more than 1000 commits) and 13,762,430 commits they created.
Eduardo Ponce, Brittany Stephenson, Suzanne Lenhart, Judy Day, and Gregory D. Peterson
ACM
The current landscape of scientific research is widely based on modeling and simulation, typically with complexity in the simulation's flow of execution and parameterization properties. Execution flows are not necessarily straightforward since they may need multiple processing tasks and iterations. Furthermore, parameter and performance studies are common approaches used to characterize a simulation, often requiring traversal of a large parameter space. High-performance computers offer practical resources at the expense of users handling the setup, submission, and management of jobs. This work presents the design of PaPaS, a portable, lightweight, and generic workflow framework for conducting parallel parameter and performance studies. Workflows are defined using parameter files based on keyword-value pairs syntax, thus removing from the user the overhead of creating complex scripts to manage the workflow. A parameter set consists of any combination of environment variables, files, partial file contents, and command line arguments. PaPaS is being developed in Python 3 with support for distributed parallelization using SSH, batch systems, and C++ MPI. The PaPaS framework will run as user processes, and can be used in single/multi-node and multi-tenant computing systems. An example simulation using the BehaviorSpace tool from NetLogo and a matrix multiply using OpenMP are presented as parameter and performance studies, respectively. The results demonstrate that the PaPaS framework offers a simple method for defining and managing parameter studies, while increasing resource utilization.
Benjamin Mayer, Joshua Arnold, Edmon Begoli, Everett Rush, Michael Drewry, Kris Brown, Eduardo Ponce, and Sudarshan Srinivas
IEEE
Reducing suicide incidence among US veterans is one of the highest priorities for the US Department of Veterans Affairs (VA). We are implementing a suicide risk detection system, in collaboration with the VA, that would serve as a surveillance system for risk factors appearing in clinical text data. Primary requirements for this system are fast search capability, feature and information extraction, and delivery of data to up-stream natural language processing models. As such, we are evaluating scalable storage solutions on the basis of performance, fault tolerance, and scalability. In this paper we present our current approach to evaluation, preliminary findings, and the work in progress towards a more robust text analysis pipeline.
Hartwig Anzt, Moritz Kreutzer, Eduardo Ponce, Gregory D Peterson, Gerhard Wellein, and Jack Dongarra
SAGE Publications
In this paper, we present an optimized GPU implementation for the induced dimension reduction algorithm. We improve data locality, combine it with an efficient sparse matrix vector kernel, and investigate the potential of overlapping computation with communication as well as the possibility of concurrent kernel execution. A comprehensive performance evaluation is conducted using a suitable performance model. The analysis reveals efficiency of up to 90%, which indicates that the implementation achieves performance close to the theoretically attainable bound.
Valerie García-Negrón, Akinola D. Oyedele, Eduardo Ponce, Orlando Rios, David P. Harper, and David J. Keffer
International Union of Crystallography (IUCr)
Composite materials possessing both crystalline and amorphous domains, when subjected to X-ray and neutron scattering, generate diffraction patterns that are often difficult to interpret. One approach is to perform atomistic simulations of a proposed structure, from which the analogous diffraction pattern can be obtained for validation. The structure can be iteratively refined until simulation and experiment agree. The practical drawback to this approach is the significant computational resources required for the simulations. In this work, an alternative approach based on a hierarchical decomposition of the radial distribution function is used to generate a physics-based model allowing rapid interpretation of scattering data. In order to demonstrate the breadth of this approach, it is applied to a series of carbon composites. The model is compared with atomistic simulation results in order to demonstrate that the contributions of the crystalline and amorphous domains, as well as their interfaces, are correctly captured. Because the model is more efficient, additional structural refinement is performed to increase the agreement of the simulation result with the experimental data. The model achieves a reduction in computational effort of six orders of magnitude relative to simulation. The model can be generally extended to other composite materials.
Juan Pablo Ponce, Eduardo Ponce, Carlos Atuesta, Tatiana Manrique, and Diego Patino
IEEE
A novel approach based on a polynomial approximation of the singular optimal control (PASOC) is proposed. The polynomial approximation solves an optimization problem for a switching system using the Hamiltonian formulation, Pon-tryagin's Minimum Principle and Legendre-Clebsch conditions, and reduces computational cost. The polynomial approximation (PASOC) is validated on a Boost converter in nominal operation, input voltage variation, and load perturbation, showing excellent performance.
Hartwig Anzt, Eduardo Ponce, Gregory D. Peterson, and Jack Dongarra
ACM
In this paper we present an optimized GPU co-design of the Induced Dimension Reduction (IDR) algorithm for solving linear systems. Starting from a baseline implementation based on the generic BLAS routines from the MAGMA software library, we apply optimizations that are based on kernel fusion and kernel overlap. Runtime experiments are used to investigate the benefit of the distinct optimization techniques for different variants of the IDR algorithm. A comparison to the reference implementation reveals that the interplay between them can succeed in cutting the overall runtime by up to about one third.
Angel González, Othoniel Rodríguez, Osvaldo Mangual, Eduardo Ponce, and Xavier Vélez
IOP Publishing
An Adaptive Embedded Digital System to perform plasma diagnostics using electrostatic probes was developed at the Plasma Engineering Laboratory at Polytechnic University of Puerto Rico. The system will replace the existing instrumentation at the Laboratory, using reconfigurable hardware to minimize the equipment and software needed to perform diagnostics. The adaptability of the design resides on the possibility of replacing the computational algorithm on the fly, allowing to use the same hardware for different probes. The system was prototyped using Very High Speed Integrated Circuits Hardware Description Language (VHDL) into an Field Programmable Gate Array (FPGA) board. The design of the Embedded Digital System includes a Zero Phase Digital Filter, a Derivative Unit, and a Computational Unit designed using the VHDL-2008 Support Library. The prototype is able to compute the Plasma Electron Temperature and Density from a Single Langmuir probe. The system was tested using real data previously acquired from a single Langmuir probe. The plasma parameters obtained from the embedded system were compared with results computed using matlab yielding excellent matching. The new embedded system operates on 4096 samples versus 500 on the previous system, and completes its computations in 26 milliseconds compared with about 15 seconds on the previous system.
G. Alvarez, Luis G. G. V. Dias da Silva, E. Ponce, and E. Dagotto
American Physical Society (APS)
A detailed description of the time-step-targeting time evolution method within the density-matrix renormalization-group algorithm is presented. The focus of this publication is on the implementation of the algorithm and its generic application. The case of one-site excitations within a Hubbard model is analyzed as a test for the algorithm, using open chains and two-leg ladder geometries. The accuracy of the procedure in the case of the recently discussed holon-doublon photo excitations of Mott insulators is also analyzed. Performance and parallelization issues are discussed. In addition, the full open-source code is provided as supplementary material.