2016 - 2017 Fellows
Electrical Engineering and Computer Sciences
Individualized Human Modeling for Medical Diagnosis and Prescription of Assistive Devices
Genetic diversity, lifestyle changes, illness, injury, age and medical intervention all affect a person’s movement, strength, and ability. Existing clinical methods are highly inefficient, requiring patients to performing a number of time consuming actions. The results of these tests are qualitative, and may not relate to the functional abilities of the individual such as self-feeding. This project aims to change the way medical diagnosis is performed by generating automated models of a persons’ abilities using affordable sensors. The results in representative models of an individual’s limbs, joints, range of motion, masses and strengths. These measurements can be tracked over time to determine recovery after injury. These measurements can also be correlated with treatments allowing the best medical intervention for a new patient to be predicted. These tools have been used in the upper limbs to quantify the severity in patients with muscular dystrophy, and the generation of assistive devices (exoskeletons).
Ruzena Bajcsy is a Professor of Electrical Engineering at UC Berkeley. Working with Robert Matthew, their research focuses on developing affordable methods for modeling a person and their abilities. This research has led to the development of individualized human models, with translation to assistive robotic devices such as exoskeletons. Their aims are to create a new framework for conducting individualized health monitoring outside of the hospital environment to reduce the risk of injury, and secondary complications and improve quality of life and recovery.
Lisa García Bedolla
Graduate School of Education
DIODE: Data Innovation for Organizing a Diverse Electorate
The United States has the lowest participation rates of any advanced industrialized country. Over the past decade, community organizations interested in making U.S. politics more participatory have adopted an integrated voter engagement (IVE) model, an approach predicated on the idea that political engagement happens most effectively through long term relationship building during and between electoral cycles. Our campaign data infrastructure, however, is not designed to provide the data support and analytics necessary to make an IVE strategy effective, particularly for groups working to engage diverse groups of voters. Through the development of a new open source data system (called DIODE) for California voters, this project will address this need, allowing organizations to maximize their civic engagement efforts and grow those efforts more effectively over time.
Professor Lisa García Bedolla is Chancellor’s Professor in UC Berkeley’s Graduate School of Education. She studies why people choose to engage politically. She has used a variety of social science methods – field observation, in-depth interviewing, survey research, field experiments, and geographic information system (GIS) – to shed light on this question. She has published four books and dozens of research articles exploring U.S. civic engagement patterns, earning five national book awards and numerous other awards. She has consulted for presidential campaigns and statewide ballot efforts and has partnered with over a dozen community organizations working to empower low-income communities of color. Through those partnerships, she has developed a set of best practices for engaging and mobilizing voters in these communities, becoming one of the nation’s foremost experts on political engagement within communities of color.
In partnership with the Social Apps Lab at CITRIS, she is working with a team of researchers that include Dr. Cristhian Parra, who obtained his Ph.D. in computer science from the University of Trento, Italy in 2009, and Rosenberg Foundation Leading Edge Fellow Michael Gómez Daly.
Electrical Engineering and Computer Sciences
VAST: High-Fidelity Network Forensics at Scale
While the past twenty years have introduced revolutionary changes in forensic techniques for criminal investigations - DNA sequencing, mass spectrometry, automated fingerprint identification - the task of analyzing cyber attacks has seen much less in the way of comparable advances. Our proposed technology, VAST (Visibility Across Space and Time), centers around developing such a capability: a platform for forensic analysis that captures and retains a high-fidelity archive of cyber-activity at the scale of an entire network, rather than a single host or network service.
Vern Paxson is a Professor of Electrical Engineering and Computer Sciences at UC Berkeley. He also leads the Networking and Security Group at the International Computer Science Institute in Berkeley, and has an appointment as a Staff Scientist at the Lawrence Berkeley National Laboratory. His research focuses heavily on measurement-based analysis of network activity and Internet attacks. He works extensively on high performance network monitoring, detection algorithms, cybercrime, and countering censorship, and co-directs the Center for Evidence-based Security Research (www.evidencebasedsecurity.org).
He will be working with a team of researchers that includes Matthias Vallentin, who will complete his Ph.D. at UC Berkeley in Spring 2016.