Mathematics

  2010 MAR 9 - (VerticalNews.com) -- "Our interest is in the problem where independent samples are drawn from two different discrete populations, possibly with a common parameter. The goal is to test hypothesis about the parameters involved in these two samples," scientists writing in the journal Statistics & Probability Letters report.

  "A number of tests are developed for the above purpose based on the Hellinger distance and penalized versions of it. The asymptotic distributions of the test statistics are derived," wrote A. Basu and colleagues, Complutense University ...read more

  2010 MAR 9 - (VerticalNews.com) -- According to a study from Beijing, People's Republic of China, "Regularized regression with the l(1) penalty is a popular approach for variable selection and coefficient estimation. For a unified treatment of thee l(1)-constrained model selection, Wang and Leng (2007) proposed the least squares approximation method (LSA) for a fixed dimension."

  "LSA makes use of a quadratic expansion of the loss function and takes full advantage of the fast Lasso algorithm in Efron et al. (2004). In this paper, we extend the fixed dimension LSA to the situation with a diverging number of parameters. We show that LSA possesses the oracle properties under appropriate conditions when the number of variables grows with the sample size. We propose a new tuning parameter selection method which achieves the oracle properties," wrote C.L. Leng and colleagues, Tsinghua University ...read more

  2010 MAR 9 - (VerticalNews.com) -- According to recent research published in the journal Statistics & Probability Letters, "The common principal components model for several groups of multivariate observations assumes equal principal axes among the groups."

  "Robust estimators can be defined replacing the sample variance by a robust dispersion measure," wrote G. Boente and colleagues.

  The researchers concluded: "This paper studies the asymptotic distribution of robust projection-pursuit estimators under a common principal components model. ...read more