XSEDE EMPOWER Positions

Please note that XSEDE EMPOWER ended with the conclusion of the XSEDE project on August 31, 2022. The following information is provided for archival purposes only.

This page provides a listing of positions that are under review for the next iteration of the program, as well as those that have been approved for current and previous iterations of the program. As part of the review and approval process, accepted student applicants are associated with an approved position.

If you are intending to mentor a student, you should create a position. If you also have a student in mind with whom you want to work, that student should submit an application. If you intend to mentor multiple students, please create a separate position for each student.

Project Title / Status	Summary	Creator / Institution
An Efficient Parallel Method for Large-Scale 3D Point Cloud Registration Completed	Point cloud registration plays an important role in many computer vision applications such as 3D object reconstruction and object detection/segmentation. However, when dealing with a large number of point clouds, many existing registration algorithms appear to have either high computational cost or limited scalability to modern parallel and distributed computing systems. In this project, we will investigate an accelerated approach to rapidly aligning multiple point clouds to a globally consistent structure by taking advantage of parallel computing. We will benchmark the proposed registration approach using real-life applications such as 3D object reconstruction with depth sensors.	Hao Ji California State Polytechnic University-Pomona Pomona, CA
Using Machine Learning for Categorizing Satellite Data (Apprentice Position) Completed	Cyclostationary signals are a special class of signals whose statistical properties do change with time in a periodic or close-to-periodic way. (https://cyclostationary.blog/) There is a lot of research that needs to be done to learn about satellite data mainly about cyclostationary signals (Gardner, 1994) and research on analyzing these data for further processing. Normally, these data needs to be manually analyzed and interesting features needs to be plotted and extracted. In this work, we want to use machine learning on the available data to extract these interesting features automatically. Unsupervised machine learning models help to analyze data and categorize data automatically based on their features. It will help group the data into various plots based on their common features. After that more work can be done including application of various other machine learning models to analyze the data such as learning about cyclostaionary data, and their extraction methods. References: Gardner, W. A. (1994). An introduction to cyclostationary signals. In Cyclostationarity in communications and signal processing (pp. 1-90). New York: IEEE press.	Sanish Rai Longwood University Farmville, VA
Using Machine Learning for Categorizing Satellite Data (Learner Position) Completed	Cyclostationary signals are a special class of signals whose statistical properties do change with time in a periodic or close-to-periodic way. (https://cyclostationary.blog/) There is a lot of research that needs to be done to learn about satellite data mainly about cyclostationary signals (Gardner, 1994) and research on analyzing these data for further processing. Normally, these data needs to be manually analyzed and interesting features needs to be plotted and extracted. In this work, we want to use machine learning on the available data to extract these interesting features automatically. Unsupervised machine learning models help to analyze data and categorize data automatically based on their features. It will help group the data into various plots based on their common features. After that more work can be done including application of various other machine learning models to analyze the data such as learning about cyclostaionary data, and their extraction methods. References: Gardner, W. A. (1994). An introduction to cyclostationary signals. In Cyclostationarity in communications and signal processing (pp. 1-90). New York: IEEE press.	Sanish Rai Longwood University Farmville, VA
RNA Comparative Analysis of COVID-19 Strains Completed	The student(s) will work with Biochem & CompSci faculty on a project aimed at understanding the patterns of the COVID19 mutations. It will include preparing the code, data capture and preparation, job configuration, job execution, and output assessments. It will also include evangelizing CI interests to one or more local high-schools.	Stephen Wheat Oral Roberts University Tulsa, OK
Large Scale Parallel Ab Initio Theoretical Techniques for Spectroscopic Study of Important Pharmaceutical Compounds Completed	The position is to support a faculty researcher in the large-scale parametric study that is estimated to require more than 25M core hours of compute capacity. The work includes developing scripts to develop the parameter space, build job descriptions, parallel job submission and management, extracting outputs from completed jobs with additional scripts to feed back into and modify the search space. Working closely with both HPC team and the Chemistry faculty researcher.	Stephen Wheat Oral Roberts University Tulsa, OK
Micro-CT Image-based Multiscale Finite Element Modeling of 3D Printed Polymeric Materials Completed	This project will conduct finite element (FE) simulations to predict mechanical properties of 3D printed polymeric materials based on their microstructures from Micro-CT images. Image segmentation and analysis will be performed on Micro-CT images to construct FE models. Realistic representative volume element (RVE) based on measurable microstructure information will then be simulated by FE analyses to predict macroscopic mechanical properties.	Jun Li University of Massachusetts-Dartmouth North Dartmouth, MA
Computational Studies of Oxidative Addition Reactions with Bidentate N-heterocyclic Carbene Nickel Complexes Completed	The apprentice will conduct quantum mechanical calculations of nickel complexes with bidentate N-heterocyclic carbene ligands. These complexes are synthesized in the PI's lab and this computational work will investigate electronic structures and mechanisms related to oxidative addition reactions of alkyl fluorides. This will complement experimental work in the area of C-F activation for applications in pharmaceutical chemistry.	S. Chantal E. Stieber California State Polytechnic University-Pomona Pomona, CA
3D Visualization, Identification, and Quantification of Biological Cells using Virtual Reality Completed	This project will involve two undergraduate students developing a novel computational tool to advance our capabilities to visualize and quantify biological cells in 3D using virtual reality. Biological data such as images of cells developing in a tissue are stored as layered 2D images gathered on instruments such as confocal microscope; this limitation in data acquisition creates pitfalls when attempting to quickly identify and quantify 3D objects such as cells that are condensed within a finite space during development or tumorigenesis. To circumvent these issues, our goal is to create an open source tool using the Unity Visualization Toolkit (UVT) that can 1) render 2D layered imaging data as distinct 3D objects in virtual reality (VR), 2) identify individual cells within a condensed space, and 3) automatically quantify cell populations.	Matthew Niepielko Kean University Union, NJ
UAV Assisted Pathfinding in Flooding Area Completed	This is follow-up research of Unmanned Aerial Vehicle (UAV) for Object Tracking. With the help of HPC, more UAV based applications can be possible. In this project, we will continue using computer vision techniques for optimal pathfinding with the assistance of UAV. UAV can capture wider view than humans. Thus, it provides better solution for a path from source to destination. This optimization problem has many real applications such as helping rescue in flooding, fire or earthquake. The students will be asked to evaluate the pathfinding algorithms in XSEDE HPC and to implement with UAV that has been built with previous funding support.	Jiang Lu University of Houston-Clear Lake Houston, TX
Regional University Learner's Cohort Completed	Learning supercomputing by preparing for the state High Performance Computing Competition, learners will experience both parallel jobs and high throughput computing.	Jeremy Evert Southwestern Oklahoma State University Weatherford, OK
Semi-supervised Medical Image Synthesis with an Average Template Completed	Fake images generated by Deep Learning models such as GAN can be visually realistic trick people's eyes. However, they are not good enough to be used to train medical staff because 1) GAN synthesize images by approximating a target distribution. This leads to a pixelized picture with good texture information but inadequate shape information. 2) Often there are not enough training data set for a specific medical feature to start with. We propose to assist the model with a pre-generated standard image which is in turn produced by morphing/merging large amounts of inputs with lesser quality.	Xinlian Liu Hood College Frederick, MD
Testing Supervised Molecular Dynamics for Curved Access Channels in p450 Completed	there are only a few hypotheses of how the drugs actually enter the enzyme in order to be metabolized. This question has implications for drug-drug interactions (effects when you need to take multiple medications) as well as drug design and discovery. This project will develop scripts to run Supervised Molecular Dynamics (SuMD) for the ingress of atorvastatin lactone (ARVL), and take preliminary data.	Edward Ackad Southern Illinois University-Edwardsville Edwardsville, IL
NSF 19-534 (Anvil Capacity Cluster) Completed	Student HPC System Administrator	Elizabett Hillery Purdue University-Main Campus West Lafayette, IN
Improving XSEDE Training Materials by Developing Exercises, Examples, and Improving Accessibility Completed	Develop short examples of common scenarios that work on several/most of the XSEDE resources. For example, develop/test/document a set of Slurm batch scripts, so a working example can be found for any XSEDE resource. Test, improve, and reorganize online training materials. Develop best practice examples for ADA accessibility in online training materials.	Susan Mehringer Cornell University Ithaca, NY
Application of Improved Dynamic Functional Connectivity Analysis Tool to Functional MRI Data Completed	Dynamic Functional Connectivity of Functional Magnetic Resonance Imaging (fMRI) data has been developed in the last decade, with the seminal methods paper by Sakoglu et al. which was applied to schizophrenia fMRI data [1-4]. A MATLAB-based software toolbox named DynaConn was developed by the PI and his former research assistant student [5]. The DynaConn has been made freely available by the PI for the neuromainging community for research [6]; however; due to lack of funding and other resources, further software development and maintainence has not been possible; thus, the toolbox's use by the has been limited. Specifically, the toolbox was developed for dynamic time-series analysis of independent component analysis (ICA) results of fMRI data. During a current XSEDE EMPOWER project in Fall 2020, the PI is supervising a senior undergraduate student to further improve the toolbox for region/atlas-based analysis, which will be nicely complementary to ICA-based analysis, and the neuroimaging community will greatly benefit from such an improved toolbox which will be offered freely by the PI. In this continuation of the project, the PI will supervise the student to apply and test extensively the improved toolbox to perform region-based dynamic functional connectivity analyses of an existing fMRI dataset that will be provided by the PI. Reference: [1] Sakoglu U, Calhoun VD, "Temporal Dynamics of Functional Network Connectivity at Rest: A Comparison of Schizophrenia Patients and Healthy Controls," Proc. 15th Annual Meeting of the Organization for Human Brain Mapping, Vol. 47, Suppl. 1, pp. S169, June 2009, San Francisco, CA. [2] Sakoglu U, Michael AM, Calhoun VD, "Classification of schizophrenia patients vs healthy controls with dynamic functional network connectivity," Proc. 15th Annual Meeting of the Organization for Human Brain Mapping, Vol. 47, Suppl. 1, pp. S57, June 2009, San Francisco, CA. [3] Sakoglu U, Calhoun VD, "Dynamic windowing reveals task-modulation of functional connectivity in schizophrenia patients vs healthy controls," Proc. 17th Annual Meeting of the International Society for Magnetic Resonance in Medicine, #3676, April 2009, Honolulu, HI. [4] Sakoglu U, Pearlson GD, Kiehl KA, Wang YM, Michael AM, Calhoun VD, "A Method for Evaluating Dynamic Functional Network Connectivity and Task-Modulation: Application to Schizophrenia," Magnetic Resonance Materials in Physics, Biology and Medicine (MAGMA), Special Issue on MR Imaging of Brain Networks, Vol. 23, pp. 351-366 (2010). [5] Esquivel J, Mete M, Sakoglu U, "DynaConn: A Software for Analyzing Brain's Dynamic Functional Connectivity from fMRI Data," Midsouth Computational Biology and Bioinformatics Society Conference (MCBIOS), March 2014, Stillwater, OK. [6] http://www.drsakoglu.com/p/dynaconndfctoolbox.html	Unal Zak Sakoglu University of Houston-Clear Lake Houston, TX
A First-principles Investigation of V2O5 as a Sensor Completed	Using density functional theory, as implemented in the VASP code, and python scripts based on the Atomic Simulation Environment (ASE), we will investigate if V2O5, a layered material, can be used as a sensor material for simple molecules, such as NH3, CO, etc. Some experiments reveal that that should be possible, and in this project we will unravel the mechanisms at an atomic level. Initial student research focused on using the quantum mechanical calculations as a black box to obtain forces and energies, which were then interpreted classically. In this continuation, we will consider the underlying calculations in more detail, and their quantum mechanical nature (e.g., wavefunctions, densities, band structures, etc.).	Hartwin Peelaers University of Kansas Lawrence, KS
Performance Tools for Holistic HPC Workflows Completed	Participate in PPerfLab research in developing performance tools for medium to large scale HPC workflows. We are developing monitoring tools to provide useful feedback to developers to guide them in addressing performance issues in their code, particularly related to workflows (applications that comprise some number of separate applications, libraries, and/or platforms) and data movement. This development requires testing various codes and gathering performance measurements with a number of different performance tools.	Karen L. Karavanic Portland State University Portland, OR
Implement Parallel Algorithm on Network Geometries Completed	We are study on the efficient implementation of efficient parallel codes for large size application problems on network geometries. Our target application problems are included power flow simulation based on TCOPF (Time constrained optimal power flow) system and traffic flow simulation which the prototype codes are already available in previous works. To implement parallel codes on network geometries, we utilized DMNetwork, Parallel Objects and Libraries of PETSc (Portable, Extensible Toolkit for Scientific Computation) of Argonne National Laboratory. Particularly, we exploit may features of PETSc DMNetwork to implement our simulation on network geometry effectively including parallel data management and parallel linear over a network geometry. We will first test the simple geometry and will extend to more complex geometry through creating functions to handle the physics on the edge of a network to extend to tackle complex geometry.	Jung-Han Kimn South Dakota State University Brookings, SD
Large-Parameter-Space DFT Calculations and MD Simulations for Fundamental PV Research Completed	The intern will assist mathematics, computer science and physics faculty and graduate students in large-scale, density functional theory (DFT) calculations and molecular dynamics (MD) simulations and visualizations. The goal is to analyze properties of Si-based and related semiconductors to aid ongoing research in photovoltaic (PV) technologies. An important role for this apprentice/intern will be to develop bash scripts to create directories for the large parameter space and associated VASP input files. Further scripting will be required to implement parallel jobs and analysis of VASP output files.	Stephen Harnish Bluffton University Bluffton, OH
Location-Aware Community Characterization Completed	Community characterization is used to assess the possibility that a given feature is over-expressed in an often complex social system. For example, the American Community Survey products provide thousands or more features for a community. In addition, a community's characteristics are also influenced by its neighboring communities. The geospatial analysis used to rely on statistics-based simulation to estimate such influence. The result is empirical without sound mathematic proof. We would like to approach the problem with the attention mechanism in deep learning to improve the understanding of the aforementioned geospatial influence in community characterization. We are originally motivated by mental health intervention, but the method could be applied to a broad range of social research topics.	Xinlian Liu Hood College Frederick, MD