Machine learning involves the development of algorithms used to
construct their own mathematical models of data to be identified
from "big data," defined as any huge amount of data that comes in
a variety of formats such as images, spectra or time-series data
that must be reviewed with in a timely manner.

This data-screening approach is now being used to identify a
variety of astronomical events from incredible amounts of data being
continuously collected by a wide-range of optical and electromagnetic
sensors, to identify events such as gravitational lensing, super novas
and exoplanets.

This approach has been developed because it is simply impossible for
human eyes to scan all available data which will include positions for
37 billion stars and galaxies as a function time. Other sensors will
collect even more data that must be screened.

Clearly, the magnitude of this review task exceeds IUSS data review
requirements.

This posting based on a article that appeared in the Dec 2017 issue of
SKY & TELESCOPE, a Harvard Observatory technical journal.

Much to be said and gained from using A.I. Facial recognition, 5G, etc.. And to think some of us still use flip phones!