Transfer Learning in Data Centers

Faster apps. More memory. Laura Brown and Zhenlin Wang bring efficiency to Big Data.

What memory resources will be available if applications A, B, and C all run together?

Big companies like Amazon and Google have even bigger data centers. Think 30 data centers each with 50,000 to 80,000 servers. And the underlying computer processors are not all identical; each year new improvements are integrated and added. Brown, Wang, and computer science colleagues from Western Michigan University are digging deep into the management of memory resources in these larger-than-life data centers.

The researchers use machine-learning techniques to create models that predict the cache and memory requirements of an application.

8

The challenge is how to make accurate predictions with such a massive variety of applications using the data center, and the different computers the application runs on. Applications might include Netflix streaming a movie, Airbnb running database queries, or NASA processing satellite images. Each app is not run in isolation with a dedicated machine. To maximize resources, data centers may have two or more applications all running on a single machine.

“If we learn the memory requirements of application A on computer X, what if the same app runs on machine Y or machine Z? Or, what memory resources will be available if A, B, and C all run together?” Brown asks.


Computational Intelligence Aids in Explosive Hazard Detection

To detect buried explosive hazards in places like Afghanistan, and to save the lives of civilians and US soldiers, Michigan Tech researcher Tim Havens realizes it requires a team—a team
of sensors.

This technology has the potential to not only save lives, but also to advance the basic science of how to combine sensors and information together to get a whole better than the sum of its parts.

A new $983,000 research project, “Heterogeneous Multisensor Buried Target Detection Using Spatiotemporal Feature Learning,” will look at how forward-looking ground-penetrating radar, LiDAR, and video sensors can be combined synergistically to see into the ground, capture high-quality images, and then automatically notify the operator of threats. With funding from the US Army Research Office, Havens and Tim Schulz, professor of electrical and computer engineering at Michigan Tech, will work with three PhD students to create a high probability-of-detection/low false-alarm rate solution.

“It’s a very difficult problem to solve because most of the radar energy bounces right off the surface of the earth,” says Havens, the William and Gloria Jackson Assistant Professor of Computer Systems at Michigan Tech. “This technology has the potential to not only save lives, but also to advance the basic science of how to combine sensors and information together to get a whole better than the sum of its parts.”

7

This new project will advance additional sensor-related work Havens and collaborators completed between 2013–2015. The US Army-funded project studied signal processing and computer-aided detection and classification using forward-looking, ground-penetrating, vehicle-mounted radar.
The Army currently fields ground-penetrating radars in its fleet. The problem is they cannot detect hazards until they’re right above them, putting a multi-million dollar radar—and soldiers—directly in the path of danger.

“The big ideas here were to process data to obtain better images, see into the ground in a high-fidelity manner, and develop algorithms to automatically find buried threats—notifying operators of what the possible threats actually are,” Havens adds.

Havens has partnered with the Army since 2008 when he was a PhD student.


New faculty joins CS this semester

image143023-persToday, we take a look at and welcome faculty who have started with the Fall Semester.

Stephane Zuckerman has joined Michigan Tech’s Computer Science Department as a visiting assistant professor. He received his PhD in Computer Sciences at the University of Versailles Saint-Wuentin_en_Yvelines (UVSQ).

Prior to coming to Tech, Zuckerman worked as a research associate and postdoctoral fellow at the University of Delaware. He has contributed to and authored numerous journal articles and presented at conferences and workshops.


Computer Science Faculty, Students Teach Kids to Code

1481146201Charles Wallace and Leo Ureel, along with two of their graduate students and six undergraduates in Computer Science, are spending time in Houghton and Hancock schools this week, giving elementary, middle and high school students hands-on experience with computer coding.

The programs are in observance of Computer Science Education Week. They include two Hour of Code events at Houghton Elementary School, one multi-day event at Houghton High School and one at Hancock Middle School. At each Hour of Code, students learn to write code, primarily using the Scratch programming language.

“We are using a tutorial developed by Michigan Tech alumna Nichole Yarroch,” Wallace said. “We are also letting students know about Computer Science and Software Engineering degrees at Tech, as well as our Copper Country Coders group that meets on the weekends.”

This is the third year that CS faculty and students have conducted Hour of Code programs at local schools.


Scott Kuhl Receives the C2E2 Award

Scott Kuhl
Scott Kuhl won an award.

Scott Kuhl received the C2E2 award in the amount of $1,600. The award will help Scott purchase affordable head-mounted displays (HMDs) to support research, education, and outreach. The Michigan Tech Century II Endowed Equipment Fund (C2E2) is a program aimed at providing equipment money to improve the lives of faculty, students, and staff campus-wide. The program is supported by the Michigan Tech Fund, through donations by individuals and organizations in the Fund.

Congratulations Scott!


Faculty and Students Attend Conference

Philart-and-StudentsMyounghoon “Philart” Jeon (CLS/CS) and his seven students attended the 8th International Conference on Automotive User Interfaces and Interactive Vehicular Applications (AutomotiveUI) Oct. 24-26 at University of Michigan.

Jeon and students hosted a tutorial on “in-vehicle auditory interactions: Design and Application of Auditory Displays, Speech, Sonification and Music.” Jeon and international collaborators hosted a workshop on “Ethically Inspired User Interfaces for Decision Making in Automated Driving.”

They had two demos at the conference: “Listen to Your Drive: An In-vehicle Sonification Prototyping Tool for Driver State and Performance Data” and “Development Tool for Rapid Evaluation of Eyes-free In-Vehicle Gesture Controls.”

This travel has been supported by CLS, CS, ICC, MTTI and HMC.


Faculty/Graduate Student Hour

Each semster Computer Science graduate students are invited to meet with faculty to share their views about the department and the graduate programs, ask questions, and discuss anything else that is of interest. It is a good time to build connections between faculty and students, and create a collaborative environment.2Faculty-Graduate Student Hour meeting photo


Tommy Stuart Receives Second Place in Elevator Pitch Competition

14606260_1272887379411026_1795585242386597791_nCongratulations to Tommy Stuart for earning second place at the 2016 Bob Mark Elevator Pitch Competition on October 6.

His pitch, “Delving Deeply,” proposed to complete development of a single-player top-down action adventure style game, and eventually start a local game development studio to leverage the large population of knowledgeable computer science students in the area.

The game idea and pitch was cultivated in Husky Game Development Enterprise in which students develop video games and were required to pitch their game ideas to the Enterprise one week before the Bob Mark competition.

For winning second place out of the 25 pitches at the Bob Mark Elevator Pitch Competition, Tommy is receiving $1,000, a free ticket to Michigan Tech’s 2017 Silicon Valley Experience trip, Smartzone Virtual Client Membership, and a Smart Start Program Tuition Waiver.


Computer Science in Top 18 in Nation

homepage_clouds_lgPayScale, a compensation analysis web site, has announced the top 25 university computer science programs in the country and Michigan Tech placed 18th.

In its 2016-2017 College Salary Report, Payscale ranked 171 colleges and universities with computer science programs based on the median early-career and mid-career pay of the schools’ computer science alumni. Tech’s early-career computer science salaries are listed at $63,900. Mid-career median pay is $126,000.

“This is great news. It is the best indicator of the quality of our programs,” said Min Song, chair of Computer Science.

Stanford University ranked number one in the nation, with its computer science graduates reporting a median early-career salary of $99,500 and mid-career salary of $168,000. Read the full report.

By Jenn Donovan