Open to all MLCD participants

• Brief report: who’s working on what
• Dataset peculiarities
  • 3.4 cosmic-string datasets are bandlimited to 0.5 Hz, although cosmic-string f_max may be higher
  • 3.5 stochastic-background datasets show non-power-law h spectrum above 0.01 Hz (synthlisa) and 0.05 Hz (lisacode). See synthlisa h spectrum
• Other questions about datasets
• Questions about challenge evaluation
• MLDC reporting at GWDAW

• Attending: Michele, Eric and Ed, Shane, Miquel, Matt, John Whelan, Ilya, Antoine, Pau, Stas, Jack Yu, Jonathan, Neil, John Baker

• Who’s working on what:
  • JPL/CIT: spinBBH search and cosmic string, will talk at GWDAW about the latter
  • APC: everything but Galaxy, considering attending GWDAW vs. LIST
  • Shane/Hellings: technique for EMRI, perhaps Galaxy, still gearing up
  • Miquel/Alberto/John Veitch: Galaxy, beginning now, will be at GWDAW
  • Reinhard/John Whelan: Galaxy, will report on training sets results at GWDAW
  • Ilya: spinBBH with JPL/CIT, TF EMRI with Jonathan, will be at GWDAW
  • AEI: Antoine on spinBBH, Stas/Jonathan on EMRIs, Stas/Krolak/Blaut on Galaxy, a subset will be at GWDAW
  • Cardiff: maximum-likelihood background study, perhaps spinBBH
  • Emma/Alberto: stochastic background
  • Montana: all the challenges

• Michele encourages abstract submissions to GWDAW; there will be a talk/panel about the future of the MLDCs
• Michele points MLDCers to his arxiv.org/0812.0751, a short review of LISA data analysis as demonstrated in the MLDCs; asks for correction of misrepresentations and missed/wrong citations

• Michele describes fmax cutoff feature of cosmic-string datasets (fmax may have been randomly chosen above Nyquist, would be equivalent to fmax = Nyquist)
• Michele describes undocumented features of stochastic background (spectrum deviates from powerlaw above 1e-2 Hz; lisacode T spectrum flattens out below 1e-3 Hz)
  • lisacode problem is due to using eccentric LISA orbits instead of rigid-rotating; plan to add a rigid-LISA dataset to the existing ones
• A note will be sent around about all this, and an analytic form for the h spectrum will be circulated
• Neil comments on more issues he sees in MLDC 3.5
  • Michele: familiar, suspected bad spectral estimation
  • Related to earlier issues in T-channel with synthlisa?

• Evaluation
  • Plan: to provide a script, in advance, for self-evaluation of entries
  • Matt’s Request:
    • Need people to do some coding/testing to make these things user friendly
      • Student’s can help:
      • First step try out the existing eval pipeline
      • Take notes on process → “user manual”
      • Can start now, but help become