In the morning, Alex Malz successfully passed his qualifying exam. He presented work on inferring the true redshift distribution given a set of noisy photometric redshifts, and proposed a thesis that extends this in various ways, including, ambitiously, to two-point functions.
In the afternoon, Andy Casey got The Cannon working with a full set of 17 labels (two stellar parameters and 15 chemical abundances). And by working, I mean training (with high signal-to-noise, well-behaved stars from APOGEE) and validation. Since this doesn't yet include our compressed-sensing regularization, Casey didn't think of this as a huge milestone, but I think that it is the highest dimensional space in which we have ever operated The Cannon. (Although I admit that we knew it would work from experiments that Melissa Ness has done.) The delivery of 15-dimensional chemical abundances for all the APOGEE stars on the red-giant branch looks feasible to me. And the labels look plausible.
No comments:
Post a Comment