The Large Synoptic Survey Telescope (LSST) will, on a nightly basis, detect and publicly release alerts on millions of “events,” sources that show measurable brightness variations relative to their historic baseline. To fully exploit this unprecedented data stream will require accurate classifications of the variables detected by LSST. The challenge, however, is that there is currently no comprehensive data set available to train automated classification algorithms for LSST. We will develop the first such data set by developing a new citizen-science, Zooniverse project. Working with Zooniverse volunteers, we will classify tens of thousands of variables, which in turn will enable fast, accurate, automated classifications by machine-learning models in the future. This project will expose LSST research and researchers to over 1 million people across the globe via the Zooniverse.
|Effective start/end date||7/1/17 → 12/31/18|
- LSST, Inc. (2017-06)