|Speaker:||Masahiro Takada (IPMU)|
|Title:||Information content in cosmological observables|
|Date (JST):||Thu, Apr 28, 2011, 13:00 - 14:30|
|Place:||Seminar Room A|
I am planning to talk about something with which you are perhaps unfamiliar.
It is about information content in cosmological datasets; how we can extract useful information from the observables, and physics/theory behind the procedures. In cosmology the two-point correlation function or its Fourier-transformed counterpart, power spectrum, is the most often-used statistics for extracting the information. The two-point correlation function is a one-dimensional quantity given as a function of the separation length. Are any other higher dimensional statistical quantities useful? For example, the most successful cosmological dataset, cosmic microwave background, is originally two-dimensional, given on the sky. Why is the one-dimenisonal power spectrum enough? What about large-scale structure probes such as galaxy surveys? I would like to talk about these things. However I haven't prepared anything for this talk yet, so may talk about totally different topics if I can't find time to prepare .... (in that case very sorry!).