Michael Franklin, UC Berkeley, Siebel Professor of Computer Science and AMPLab Director
The Power of Data-Driven Science
- THE EXPLOSION OF DATA IS
THE RESULT OF SO MUCH
INFORMATION NOW BEING
DIGITIZED. - EASY ACCESS TO DATA AND
LOW-COST ANALYTICAL
TOOLS HAVE BECOME
THE BASIS FOR A NEW
SCIENTIFIC PARADIGM:
DATA-DRIVEN SCIENCE.
“The amount of data is doubling every two years, and that rate of data accumulation is accelerating.”
Data have always played a central role in scientific research. Historically, good data were difficult to obtain, so scientific methods had been used to identify and generate meaningful data. For instance, an experimental approach to science involves devising a controlled experiment, usually with a hypothesis in mind, and then observing the results. The experimental data either prove or disprove the hypothesis. A theoretical approach to science involves developing mathematical explanations of
phenomena, and then over time collecting data that support or contradict the formula. A computational approach to science involves developing computer models of complex phenomena, such as weather or models of social behavior, and then comparing predictions to actual outcomes. All of these methods depend on limited data sets.
This is an excerpt from Data Disruption. The eBook was generously sponsored by Mighty Guides.
