Skip to content

Feb. 15, 2019

Jump to bottom
David Lawrence edited this page Feb 15, 2019 · 6 revisions

Location: CC F326-327

Time: 14:30-15:30

Agenda

  1. Description(David)
  2. Toy Problem (Gagik)
  3. Meeting time and frequency
  4. AOB

Minutes

Attendees: Simon T., David L., Gagik G., Markus D., Thomas B., Ole H.

  • We discussed the toy problem (ex_toy2).
  • Both David and Gagik showed slides describing networks they trained and the resulting angular resolutions.
  • David's network (MLP) resulted in a ~16 degree resolution
  • Gagik's network was able to achieve 0.3 degree resolution.
  • There was some discussion on the reason for the discrepancy. It was speculated that this was due to David's samples having a variation on the origin of the track whereas Gagik's sample had all tracks originating from the center. (The speculation was confirmed after the meeting).
  • We discussed the next challenge. Gagik suggested fitting arcs drawn from the center. The radius of curvature (p) and phi angle can both be fit and two values extracted.
  • David volunteered to set up a JLab mailing list for communications.

Next meeting: We agreed to meet about every 2 weeks. The next meeting will be on Friday Mar. 1st at 2:30pm

Clone this wiki locally