Olga edited Methods Description.tex  over 9 years ago

Commit id: f8f4cf3b94ebac2088b93817eb6a9902098575f6

deletions | additions      

       

\section{Methods Description} \label{sec:methods}  \subsection{GMM} \label{sec:methods-gmm}  To Olga and Yingyi, you guys should fill in this part of the article.  \href{http://en.wikipedia.org/wiki/Expectation\%E2\%80\%93maximization_algorithm#mediaviewer/File:EM_Clustering_of_Old_Faithful_data.gif}{GMM (Gaussian mixture modeling)}  methods maximize method maximizes  the liklihood likelihood  of the data set using EM (expectation-maximization) method method.  Input data:  $K = 2$ (number of mixture components)  $N = $ number of observations  $\theta = (\mu_{1},\mu_{2}, \sigma_1, \sigma_2, p)$  \subsection{Excess-mass method (EMM)} \label{sec:methods-emm}