SVM's, score-spaces and maximum margin statistical models Export

@unpublished{citeulike:568501, abstract = {There has been significant interest in developing new forms of acoustic model, in particular models which allow additional dependencies to be represented than allowed within a standard hidden Markov model (HMM). This paper discusses one such class of models, augmented statistical models. Here a locally exponential approximation is made about some point on a base distribution. This allows additional dependencies within the data to be modelled than are represented in the base distribution. Augmented models based on Gaussian mixture models (GMMs) and HMMs are briefly described. These augmented models are then related to generative kernels, one approach used for allowing support vector machines (SVMs) to be applied to variable length data. The training of augmented statistical models within an SVM, generative kernel, framework is then discussed. This may be viewed as using maximum margin training to estimate statistical models. Augmented Gaussian mixture models are then evaluated using rescoring on a large vocabulary speech recognition task.}, author = {Gales and Layton}, citeulike-article-id = {568501}, comment = {For every point of the family, finds a tangent exponential family and throws it into the mix, resuling in "augmented statistical models", also slides from NIPS 2005 workshop}, keywords = {exponential-families, generative-discriminative, svm}, posted-at = {2006-03-29 20:13:52}, priority = {2}, title = {SVM's, score-spaces and maximum margin statistical models}, year = {2005} }

TY - UNPB ID - citeulike:568501 L3 - citeulike-article-id:568501 N2 - There has been significant interest in developing new forms of acoustic model, in particular models which allow additional dependencies to be represented than allowed within a standard hidden Markov model (HMM). This paper discusses one such class of models, augmented statistical models. Here a locally exponential approximation is made about some point on a base distribution. This allows additional dependencies within the data to be modelled than are represented in the base distribution. Augmented models based on Gaussian mixture models (GMMs) and HMMs are briefly described. These augmented models are then related to generative kernels, one approach used for allowing support vector machines (SVMs) to be applied to variable length data. The training of augmented statistical models within an SVM, generative kernel, framework is then discussed. This may be viewed as using maximum margin training to estimate statistical models. Augmented Gaussian mixture models are then evaluated using rescoring on a large vocabulary speech recognition task. TI - SVM's, score-spaces and maximum margin statistical models KW - exponential-families KW - generative-discriminative KW - svm N1 - For every point of the family, finds a tangent exponential family and throws it into the mix, resuling in "augmented statistical models", also slides from NIPS 2005 workshop AU - Gales AU - Layton PY - 2005/// ER -