[HydraTeam] [EXTERNAL] Poster from Meeting

[David Lawrence]

Hi Manav,

  I took another look at the poster. Overall, it looks good. Here are a few comments/questions.

1. The pie chart on left shows “No Data” as a categories and a number “21” looks to be indicating
that at 1 of 4 white “spokes” in the plot.

2. The two methods look to be presented quantitatively in different ways. A plot for Hydra and a
table for the Non-ML method. Given the bullet about the 1275 plots, I’m not sure I know how to
compare the two. Does it make sense to add a horizontal line to the plot? Hydra numbers to the
table?

I see two significant results here:

A. Hydra model takes relatively little data to generalize the problem of identifying “Bad"

B. A custom algorithm designed specifically to classify a known problem doesn’t necessarily
do better than an ML-model for a broad classification. They *can* though be coupled to
refine the classification in cases where it doesn’t make sense to train a model (i.e. class for
every possible bad module).

I understand there is a tight timeline so take these as minor comments on the poster that
don’t necessarily require action. I would, however, suggest you make sure to emphasize
these point in the thesis.


Regards,
-David

-------------------------------------------------------------
David Lawrence Ph.D.
Staff Scientist - - EPSCI Group Lead
Thomas Jefferson National Accelerator Facility
Newport News, VA
davidl at jlab.org
(757) 269-5567 W
(757) 746-6697 C

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/hydrateam/attachments/20240307/aa923e5c/attachment.html>