Abstract:

There are several analysis problems in astronomical data that appear to be obvious or easy, but lack rigorous statistical solutions. I will discuss several of these problems, such as recognizing the boundary of irregular objects, sifting the data into spatially, spectrally, and temporally coherent segments, detecting sharp and fleeting transients, understanding the effects of wellknown but unmodeled systematics, recognizing the domain where an observable is distributed as a powerlaw, computing uncertainties on computationally expensive simulations, etc. I will highlight some of the advances made by our CHASC Astrostatistics Collaboration.
