Physical Sciences and Mathematics Commons^™

Constrained K-Means Clustering Validation Study, Nicholas Mcdaniel, Stephen Burgess, Jeremy Evert

Student Research

Machine Learning (ML) is a growing topic within Computer Science with applications in many fields. One open problem in ML is data separation, or data clustering. Our project is a validation study of, “Constrained K-means Clustering with Background Knowledge" by Wagstaff et. al. Our data validates the finding by Wagstaff et. al., which shows that a modified k-means clustering approach can outperform more general unsupervised learning algorithms when some domain information about the problem is available. Our data suggests that k-means clustering augmented with domain information can be a time efficient means for segmenting data sets. Our validation study focused …

Validation Study Of Image Recognition Algorithms, Jacob Miller, Jeremy Evert

Student Research

Developments in machine learning in recent years have created opportunities that previously never existed. One such field with an explosion of opportunity is image recognition, also known as computer vision; the process in which a machine analyzes a digital image.

In order for a machine to ‘see’ as a human does, it must break down the image in a process called image segmentation. The way the machine goes about doing this is important, and many algorithms exist to determine just how a machine will decide to group the pixels in an image.

This research is a validation study of related …

Profiling The National Center For Atmospheric Research’S Weather Research And Forecasting Model, Jacob Miller, Jeremy Evert

Student Research

This project employs High Performance Computing (HPC) to analyze the National Center for atmospheric research’s Weather, Research, and Forecasting Model (NCAR-WRF). Once an appropriate profiler is found, it will be used to determine if NCAR-WRF is compute bound, meaning its speed is determined by the power of its processors, or memory bound, meaning that the wall clock time is determined by the amount of available memory.

Introducing Hpc To Local High School Curriculums, Madison Matli, Devin Smoot, Jeremy Evert

Student Research

High Performance Computing is a topic that hasn’t been introduced or discussed in most Oklahoma high school curriculums. While students are being pushed to take more general electives, they are missing out on important knowledge pertaining to their field of study. This process usually occurs because students lack access to educational materials or because the school simply does not offer certain specialty courses. Learning high level information at a young age can help students excel in future college courses and research projects. My research is focused on providing Oklahoma high schools with a simplistic High-Performance Computing guide that walks through …

Perfsonar And Offn: Diagnosing And Testing Science Dmz Connections For The High-Speed Transfer Of Large Datasets On Oneoklahoma Friction Free Network, Dylan Ortega, Jeremy Evert

Student Research

The OneOklahoma Friction Free Network (OFFN) is a Science DMZ network that is designed with the purpose of enabling collaboration between high performance computing centers and higher learning institutions from across the state of Oklahoma. OFFN allows researchers the ability to transfer datasets at speeds faster than a standard internet connection by allowing direct connections to servers on OFFN by not having to deal with firewalls. We use PerfSonar to measure the performance of OFFN. A PerfSonar testing server is needed on each OFFN node to monitor the health of the network as a whole. PerfSonar can perform response testing, …

A Literature Review Of Research On The Cosmic Microwave Background, Hayden Webb, Jeremy Evert

Student Research

The objective of this research is to establish a better understanding of the current work available in the field of the Cosmic Microwave Background (CMB).

Our thesis is that, while the current state of research has created a base for the understanding of the CMB, more work is necessary through the use of computer science methods to expand upon the field. Our method of study is a literature review.

Our preliminary work suggests that NASA holds a wealth of information on this topic and has made it freely available to the public. Through our partnership with the OneOklahomaCyberinfrastructure Initiative (OneOCII), …