Документ взят из кэша поисковой машины. Адрес оригинального документа : http://hea-www.harvard.edu/AstroStat/slog/groundtruth.info/AstroStat/slog/2007/arxiv-matching-sources/feed/index.html Дата изменения: Unknown Дата индексирования: Mon Feb 4 06:18:19 2013 Кодировка: Поисковые слова: rainbow

http://hea-www.harvard.edu/AstroStat/slog/2007/arxiv-matching-sources/ Weaving together Astronomy+Statistics+Computer Science+Engineering+Intrumentation, far beyond the growing borders Fri, 01 Jun 2012 18:47:52 +0000 hourly 1 http://wordpress.org/?v=3.4 http://hea-www.harvard.edu/AstroStat/slog/2007/arxiv-matching-sources/comment-page-1/#comment-163 hlee Tue, 19 Feb 2008 04:51:20 +0000 http://hea-www.harvard.edu/AstroStat/slog/2007/arxiv-matching-sources-july-11-2007/#comment-163 The draft update was told a while ago but I've been lazy to review both my posting and the update. Although I cannot comment on multiple testing for source matching and what are the actual changes from the first draft to the latest one, I can say that my doubt on penalizing complicated hypotheses over simpler ones is cleared. Yet, another question remains on the curse of dimensionality, independent of the paper. Not knowing cosmology (mass distribution) and computation for two hypotheses (H: all matching vs K:at least one is not matching), I wonder how Bayes factor will be executed (I saw a statement as follows: BF(H,K) larger than 10 strongly favors H; 2-10 mildly favors H, etc.). Are there any power related studies on source matching statistics? Although I don't comprehend the history of astrometric source matching, I see the work by Budavari and Szalay will lay a corner stone to develop feasible and practical source matching statistics. The draft update was told a while ago but I’ve been lazy to review both my posting and the update. Although I cannot comment on multiple testing for source matching and what are the actual changes from the first draft to the latest one, I can say that my doubt on penalizing complicated hypotheses over simpler ones is cleared. Yet, another question remains on the curse of dimensionality, independent of the paper. Not knowing cosmology (mass distribution) and computation for two hypotheses (H: all matching vs K:at least one is not matching), I wonder how Bayes factor will be executed (I saw a statement as follows: BF(H,K) larger than 10 strongly favors H; 2-10 mildly favors H, etc.). Are there any power related studies on source matching statistics? Although I don’t comprehend the history of astrometric source matching, I see the work by Budavari and Szalay will lay a corner stone to develop feasible and practical source matching statistics.

]]>