Difference: 2018-03-08 ( vs. 1)

ZTF ML Minutes - 20180308

Attendees: Umaa, Brian, Ashish, Richard, Matthew, Frank, Charlotte, Nadia, ..?

ZTF update (Matthew): - Weather affecting recent observations. i-filter may not be in until later.

Pipeline updates (Frank): - AVRO packets now have more metrics based on the discussions earlier, and the handling capabilities. - S/G scores still being updated - cutouts being changed to fits

Zooniverse update( Richard): - currently i-band set is in. Will be switched out to include sources with RB < 0.1, RB > 0.9, 0.4 < RB < 0.6

RB-status (Umaa): - duplicate classifications from zooniverse. While that is desirable when we have many scanners, for now it will be restricted to 1. - bogus from transient marshall downloaded nightly. Not yet incorporated in model - (slides attached)

Deep Learning status (Brian): - experiments on iPTF data are promising - need more time to run, but should provide better results.

RF Tree interpreter (Charlotte working with Umaa): - see attached slides.

Overall (Ashish): - we will be training some students to run some of the ML pipeline as we automate it.

-- AshishMahabal - 17 Mar 2018