Recently developed adaptive Markov chain Monte Carlo (MCMC) methods have been applied successfully to many problems in Bayesian statistics. Grapham is a new open source implementat...
tion Abstract ChengXiang Zhai (Advisor: John Lafferty) Language Technologies Institute School of Computer Science Carnegie Mellon University With the dramatic increase in online in...
This paper explores two classes of model adaptation methods for Web search ranking: Model Interpolation and error-driven learning approaches based on a boosting algorithm. The res...
Jianfeng Gao, Qiang Wu, Chris Burges, Krysta Marie...
Extracting information from data, often also called data analysis, is an important scienti c task. Statistical approaches, which use methods from probability theory and numerical a...
Bernd Fischer 0002, Johann Schumann, Thomas Pressb...
A quaternion widely linear (QWL) model for quaternion valued mean-squared-error (MSE) estimation is proposed. The augmented statistics are first introduced into the field of quater...