Reading Consumers' Minds: Q&A with Prof. Yoav Freund
The San Diego Union-Tribune got wind of a report that Amazon patented technology to greatly improve the company's ability to quickly ship products by sending them toward consumers before they had even placed an order. To find out what this "anticipatory shipping" would mean, the newspaper turned to CSE Prof. Yoav Freund (at right) for a Q&A that appeared on Feb. 27. Freund began by clarifying that the patent was not for some magical way to send a product before you order it. "What Amazon patented is a very specific tweak: instead of using storage in a store or storage room close to the customer, you use the UPS or Fedex distribution point," explained Freund. "You send the package to the UPS office, and if you get a matching order when the package is in transit, you add the address to the package and let it continue directly to the customer. This will require a special arrangement with UPS/Fedex but it does not require making better predictions. In other words, it changes the economics of the supply chain, but the principle is the same."
The paper went on to ask whether Freund worries about the speed with which sicence is figuring out how to monitor and track people. The CSE professor said his main worry in this regards was ot about standard commerce. "I am much more worried about the implications on healthcare, on insurance and on privacy in general," he said, adding that "analysis of health-related information has a great potential of improving healthcare and making it cheaper. However, when healthcare is a business, its interests might not always align with those of the patient, and without the proper tools and regulation in place, there is a great potential for abuse." Freund noted that there is a lot of research on how to do machine learning and statistics in a privacy-preserving fashion. Read the full Q&A at UTSanDiego.com.
CSE Affiliated Professor Provides Key Technology for Consumer Robot
While UC San Diego pursues plan to create a new multidisciplinary robotics institute on campus, it's clear that Thomas Bewley (photo at right by UT's Howard Lipin) will be a key player. The professor of mechanical and aerospace engineering is also an affiliated member of the CSE faculty. As reported in the San Diego Union-Tribune on Jan. 27, Bewley "developed much of the technology that powers, balances and controls MiP," which stands for "mobile inverted pendulum," a toy robot to hit U.S. shelves in May. The robot comes from WowWee, a company with U.S. operations based in La Jolla. According to the Union-Tribune, the toy is 7.5 inches tall, weighs 12.3 ounces, and is a "two-wheel, self-balancing robot that can be controlled with hand gestures or commands from smartphones. It rolls at a quick clip, it can dance to scores of songs and it plays a variety of games." The device is also hackable, and WowWee U.S. president Peter Yanofsky is quoted as saying, "This will get kids into science." According to campus officials, Bewley's technology licensed for MiP could earn as much as $1 million in technology transfer fees, and even more from future iterations of MiP and other robots on the drawing-board at WowWee and UCSD. Read the full story at UCSanDiego.com.
Crowdsourcing a Living Map of World Health
What if by collecting data from mobile medical apps on cell phones around the world, we could map significant problems and see the flu coming like a giant whirling hurricane? A team of computer scientists, engineers, biologists and medical researchers at UC San Diego wants to leverage the widespread use of smart phone technology and cloud computing to build maps of large-scale health problems or environmental damage such as the concentration of heavy metals in drinking water. The idea is based on the principle that health, including infectious disease and environmental pollution, is a trackable geospatial event. The team is working towards developing a tricorder (see below) that could monitor both individual and environmental health. In phase one, citizen sensors will test their drinking water using a simple test strip device that automatically sends the test results to a central data server for analysis while telling the tester whether the water is safe to drink.
The Jacobs School of Engineering's Catherine Hockmuth sat down with Andrew Huynh (see above), a CSE Ph.D. student who is leading development of the team's data storage and analysis platform, to discuss the role of machine learning in the project. In machine learning, computers learn how to do something from the continuous input and analysis of data. Huynh says the picture can become hazy with inaccurate information when humans are in the loop. Part of his job is to train the computer to learn to distinguish accurate data from erroneous data.
Q: What is your role as lead data scientist on this project?
A: As lead data scientist I helped architect and develop our open health stack, a cloud analytics and storage platform that will be used to detect large-scale trends in data from sensors, individuals, and the environment. I also manage a small team of undergraduate researchers who work on various components as a way to gain knowledge of the latest best practices in programming.
Ph.D. Students to Showcase Research Findings from Fall Florence Expedition
On Feb. 20, CSE graduate students who are part of the Qualcomm Institute's Center of Interdisciplinary Science for Art, Architecture and Archaeology (CISA3) will take part in the CISA3 Open Lab Night. Specifically, they will be showing some of their research findings from their field expeditions in Fall 2013 that took most of them to Greece, France and Italy. CSE Ph.D. students Vid Petrovic, David Vanoni, and John Mangan spent most of the quarter in Florence working on their individual research projects. (Pictured l-r: Expedition leader Maurizio Seracini, archaeology student Ashley Richter, structural engineering student Mike Hess, CSE students Mangan, Vanoni and Petrovic, visiting scholar Caterina Maffei, and Duomo architect.) But they also took part in multidisciplinary projects with CISA3 colleagues from cognitive science, materials science and engineering, and structural engineering. They were given exclusive access to document one of the oldest buildings in Florence, the Opera del Duomo cathedral, including the Baptistery of Saint John, Brunelleschi's Duomo, Giotto's Bell Tower, and the crypt dating to the site's Roman origins.
Working in early mornings and late at night, the researchers undertook an ambitious program of imaging and laser scanning. The buildings were captured via terrestrial laser scanning, thermography, high-resolution photogrammetry (including structure-from- motion), and stereoscopic (3D) imaging. In addition, they undertook multispectral imaging of both the interior and exterior of the building, emphasizing spaces to which the public has no access. The ultimate goal was to produce information that will be consolidated into what CISA Director Falko Kuester calls "digital, annotated scaffolds for data." CISA3 also signed a five-year memorandum of understanding to continue work closely with the Opera del Duomo. The researchers also continued a multi-year campaign documenting the status of the Palazzo Vecchio. In previous years, most of the focus was on the Great Hall and the search for a lost Leonardo da Vinci mural. This year CISA3 students acquired a point-cloud model of the entire exterior of the historic structure. "We did all o f the exterior laser scanning late at night," said CSE Ph.D. student John Mangan, who is researching techniques for cloud-centric data processing and communication of rich cultural heritage data collections "That's when the Piazza della Signoria and surrounding streets were at their emptiest." CSE Ph.D. student David Srour was on site, investigating point-based data representation strategies, enabling the synthesis, visualization and intuitive exploration of the collected data.
"We extended our laser scanning, high-resolution imaging, and multispectral imaging survey to include other rooms not previously surveyed by CISA3 on the far side of Palazzo Vecchio," noted CSE student Vid Petrovic(pictured at right). "The main emphasis was on the Hall of Geographical Maps and surrounding spaces." For Petrovic, the field expedition provided a one-on-a-kind living laboratory for the evaluation, application and refinement of rapid Big Data visualization techniques, allowing billions of data points to be synthesized and visually explored on the fly, informing the work conducted by the highly interdisciplinary CISA3 team. By taking research from the lab into the "the wild," CISA3 students gain important experience, going from theory to practice while working with domain experts solving real-world problems. Working in the Hall of Geographical Maps gave the students access to the Mappus Mondi, a central globe that is now unreadable. Says CSE student David Vanoni: "We are now looking into whether non-intrusive imaging methods can reveal what is on the map and how we can make that information available to the touring public." Read the full news release.
January 29, 2014 - 11am-Noon - Room 4140, CSE Building
University of Washington professor Michael Ernst will talk about the "Collaborative Verification of Information Flow for a High-Assurance App Store." He will explore a vision of a verified app store in which each application has been formally proven to be free of certain defects and exploits.
January 31, 2014 - 11am-Noon - Room 1202, CSE Building
Princeton University computer science lecturer Josh Hug will explore "MOOC and Me: Development of the Princeton Algorithms MOOC and its On-Campus Counterpart." Hug will discuss his experience building a MOOC as well as a flipped lecture experiment conducted in Spring 2013, as well as plans to extend the experiment in Spring 2014.
February 3, 2014 - 11am-Noon - Room 1202, CSE Building
UC Berkeley professor Ravi Ramamoorthi is the latest speaker in the CSE Distinguished Lecture Series. He will talk about "Sampling and Reconstruction of High-Dimensional Visual Appearance," including results which point to a unified sampling theory applicable to many areas of signal processing, computer graphics and computer vision.
February 5, 2014 - 11am-Noon - Room 1202, CSE Building
McGill University lecturer Anil Ada will present on "Communication Complexity." The topic includes a study of privacy in the context of communication complexity: how much information do the players reveal about their input when following a communication protocol.
February 10, 2014 - 2-3pm - Room 4140, CSE Building
Microsoft Research staff researcher Alexandr Andoni delivers a Theory and Communications Sciences Seminar.
The "Big Data Big Network 2" workshop is the second international conference on Mexican-American research collaboration partnerships, and provide information and training on how best to use high-speed networking and display walls. SDSC distinguished scientist Chaitan Baru will deliver the keynote on "Big Data and Fat Pipes: Building a Shared Infrastructure for Data Research at Scale."
This seminar by IDC Herzliya professor of communication Guy Hoffman is the second in the Design at Large series for Winter 2014 organized by CSE Prof. Scott Klemmer. Hoffman will talk about "Human-Robot Teamwork: Fluency and Embodiment in Artificial Intelligence."
UC San Diego Extension and UC-TV are organizing a conference on "UCSD Big Data at Work," primarily to serve the campus and San Diego communities with information about the Big Data challenge to existing industries, and the most valuable future jobs in the Big Data sector. Speakers on the program include SDSC director Mike Norman, Calit2 director Larry Smarr, CSE Prof. Stefan Savage, among others.
Have a notice about upcoming travel to conferences, etc., for the Faculty GPS column in our weekly CSE Newsletter? Be sure to let us know! Email Doug Ramsey at email@example.com.