In an interview, University of California, San Diego (UCSD) Ph.D. student Andrew Huynh discussed the role of machine learning in a project that uses crowdsourcing to create maps of large-scale health problems and environmental damage.
UCSD and the Qualcomm Institute are developing a tricorder that could monitor both individual and environmental health. Using smartphones and cloud technology, citizens will transmit data such as the concentration of heavy metals in drinking water to a central server for analysis.
Huynh serves as lead data scientist on the project, and helped develop the cloud analytics and storage platform that will be used to identify large-scale trends in data from sensors, individuals, and the environment. Using machine learning, Huynh is training the computer to differentiate accurate from inaccurate data. He notes that although the research value of crowdsourcing is increasingly recognized, data quality is an important issue to address. Data are entered "by people who are prone to accidents or misperception rather than a deterministic machine," Huynh says. "Determining the signal from the noise in these huge datasets is a massive problem." When the project's sensors are in place, Huynh and his team will experiment with different methods of discerning data quality.
From UCSD News (CA)
View Full Article
Abstracts Copyright © 2014 Information Inc., Bethesda, Maryland, USA
No entries found