The University of Winsconsin-Madison has established a new institute that will focus on advancing big data science, a leading technology trend in modern times. Big data has become a catchphrase that refers to the area of computer science which deals with huge troves of information.
Over the last two decades, almost every academic field and industry has changed, thanks to massive datasets. In astronomy, researchers are now using big data tools in analyzing data from telescopes and gather new insights about the stars. Politicians have also used big data analytics during campaigns to target voters more effectively.
The Institute for Foundations in Data Science will be part of the Wisconsin Institute for Discovery. Its focus will be the re-examination of the core statistics, computer science and mathematics powering big data science with the aim of coming up with new ways of using large datasets more effectively and efficiently.
According to Stephen Wright, a professor of computer science and the leader of the 14-member faculty behind the initiative, big data faces several looming challenges that make it necessary to revisit the fundamentals. More challenges are expected as big data gets even bigger. The fact that the size of datasets keeps on exploding makes it difficult to develop computational frameworks and algorithms to keep up with the scale.
Wright points out that current methodologies are struggling with many aspects of large datasets. As a result, a lot of data is noisy, fuzzy or is missing. Moreover, the limits on how powerful computers can become as demonstrated by the slowing down of advancements in computer processor developments over the last five years is a reason to be concerned.
Wright says that data analysis and processing are very intensive tasks. It is therefore important to be concerned that computers are no longer improving as rapidly as they should be.
The institute will be taught by UW faculty members who have excelled in data science innovations. One of them will be Wright. Others include Robert Nowak, a professor of engineering who used big data to come up with an app capable of mapping out flavours of beer therefore accurately predicting a person’s preference. It works like Pandora which predicts music preferences. The faculty will also be privileged to have professor Sebastian Roch, a mathematician who has successfully mapped out complex patterns of evolution among species thus making a splash in “computational biology.” The faculty will also have Rebecca Willet who is an associate professor of computer and electrical engineering and has done a lot in medical imaging.
Wright has disclosed that the institute will focus on three areas of research as part of the first phase of its work. These include:
- Expressing big data problems as mathematical statements
- Graphing big data problems
- Figuring out better ways of collecting data.
The group has already received some grant from the National Science Foundation (NSF) to start its operations. Wright hopes that the institute will make good progress which will lead to more funding from the NSF to eventually grow into a larger enterprise capable of working with other groups across the work on data science innovation.