NSF REU students contribute to, learn from Aristotle Cloud Federation project
Contact: Paul Redfern
Cell: (607) 227-1865
FOR RELEASE: March 28, 2018
ITHACA, NY – Five National Science Foundation Research Experiences for Undergraduates (REU) students made valuable contributions to Aristotle Cloud Federation science use cases at Cornell University and the University of California, Santa Barbara, and gained domain-specific knowledge and first-hand experience on how to use cloud systems effectively for data analysis.
In support of Cornell professor Sara C. Pryor’s atmospheric science research, REU student Thomas Biondi:
- processed over 10 terabytes of complex datasets produced by weather models on multiple computing platforms
- used Aristotle to store the large amounts of data while simultaneously running the data through code on a Linux virtual machine
- applied statistical metrics to compute weather model accuracy
- used machine learning algorithms to predict days with high wind speeds
- created visualizations to present the data in a way that a lay audience could understand.
Also, at Cornell, working with recent PhD recipient Robert Wharton and Aristotle science team lead Adam Brazier, REU students Elizabeth Holzknecht and Shiva Laskhamann built code to down-resolve radio astronomy search data, produced graphical output, and rebuilt code for production runs and extension to new algorithms. Their code will be run on datasets known to contain transient sources and on pulsar blind search datasets being investigated by Cornell astronomy professor James Cordes and his colleagues.
Cornell entomology professor Angela Douglas’s REU student, Joan Song, enabled initial research on three-compartment microbial models, focusing on the bacterial community in the fruit fly Drosophila. The Douglas Lab investigates the interactions between animals and beneficial microbes as a biomedical model for human health and a novel target for insect pest control.
At University of California, Santa Barbara, REU student William Berman has been researching the ability to predict market prices in the Amazon Web Services (AWS) spot market for cloud resources. Working with Durability Agreements from Time Series (DrAFTS) – an on-demand service for generating reliability guarantees in the AWS spot market developed by professor Rich Wolski – he developed a new web-service venue that is easier to deploy, more scalable, and has a more user-friendly interface than the original prototype. He is also looking at using DrAFTS to implement a long-lived virtual machine service with the same longevity characteristics as the on-demand service, but at a significantly lower price.
The Aristotle Cloud Federation project – https://federatedcloud.org/ – is supported by National Science Foundation grant ACI-1541215 and the Cornell University Center for Advanced Computing (CAC) in partnership with the University at Buffalo Center for Computational Research (CCR) and the University of California, Santa Barbara Department of Computer Science.
The Research Experiences for Undergraduates (REU) program – https://www.nsf.gov/crssprgm/reu/ – supports active research participation by undergraduate students in any of the areas of research funded by the National Science Foundation.