A View of the EM Algorithm that Justifies Incremental, Sparse, and other Variants Export

@inproceedings{neal-hinton-98, abstract = {. The EM algorithm performs maximum likelihood estimation for data in which some variables are unobserved. We present a function that resembles negative free energy and show that the M step maximizes this function with respect to the model parameters and the E step maximizes it with respect to the distribution over the unobserved variables. From this perspective, it is easy to justify an incremental variant of the EM algorithm in which the distribution for only one of the unobserved variables...}, author = {Neal, R. and Hinton, G.}, booktitle = {Learning in Graphical Models}, citeulike-article-id = {686689}, citeulike-linkout-0 = {http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.33.2557}, editor = {Jordan, M. I.}, keywords = {clustering, gaussian, learning}, posted-at = {2008-06-02 07:20:59}, priority = {2}, publisher = {Kluwer}, title = {A View of the EM Algorithm that Justifies Incremental, Sparse, and other Variants}, url = {http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.33.2557}, year = {1998} }

TY - CONF ID - neal-hinton-98 L3 - citeulike-article-id:686689 N2 - . The EM algorithm performs maximum likelihood estimation for data in which some variables are unobserved. We present a function that resembles negative free energy and show that the M step maximizes this function with respect to the model parameters and the E step maximizes it with respect to the distribution over the unobserved variables. From this perspective, it is easy to justify an incremental variant of the EM algorithm in which the distribution for only one of the unobserved variables... TI - A View of the EM Algorithm that Justifies Incremental, Sparse, and other Variants PB - Kluwer T2 - Learning in Graphical Models KW - clustering KW - gaussian KW - learning AU - Neal, R AU - Hinton, G A2 - Jordan, MI PY - 1998/// UR - http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.33.2557 ER -