<\/p>\n\n\n
Deep neural networks achieve state-of-the-art results in various classification and synthetic data generation tasks. However, only little is known about why depth improves a model. We investigate the structure of stochastic deep neural works, also known as Deep Boltzmann Machines, to shed some light on this issue. While the best known results postulate an exponential dependence between the number of visible units and the depth of the model, we show that the required depth is upper bounded by the longest path in the underlying junction tree, which is at most linear<\/em> in the number of visible units. Moreover, we show that the conditional independence structure of any categorical Deep Boltzmann Machine contains a sub-tree that allows the consistent estimation of the full joint probability mass function of all visible units. We connect our results to L1-regularized maximum-likelihood estimation and Chow-Liu trees. Based on our theoretical findings, we present a new tractable version of Deep Boltzmann Machines, namely the Deep Boltzmann Tree<\/em> (DBT). We provide a hyper-parameter-free algorithm for learning the DBT from data, and propose a new initialization method to enforce convergence to good solutions. Our findings provide some theoretical evidence for why a deep model might be beneficial. Experimental results on benchmark data show, that the DBT is a theoretical sound alternative to likelihood-free generative models.<\/p> <\/div>\n<\/div>\n\n\t <\/div>\n\t \t<\/div>\n\t\t <\/div>\n\t \n\t<\/div>\n\t\n\t<\/div>\n\n","protected":false},"excerpt":{"rendered":" Deep neural networks achieve state-of-the-art results in various classification and synthetic data generation tasks. However, only little is known about why depth improves a model. We investigate the structure of stochastic deep neural works, also known as Deep Boltzmann Machines, to shed some light on this issue. While the best known results postulate an exponential […]<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"builder_content":" Deep neural networks achieve state-of-the-art results in various classification and synthetic data generation tasks. However, only little is known about why depth improves a model. We investigate the structure of stochastic deep neural works, also known as Deep Boltzmann Machines, to shed some light on this issue. While the best known results postulate an exponential dependence between the number of visible units and the depth of the model, we show that the required depth is upper bounded by the longest path in the underlying junction tree, which is at most linear<\/em> in the number of visible units. Moreover, we show that the conditional independence structure of any categorical Deep Boltzmann Machine contains a sub-tree that allows the consistent estimation of the full joint probability mass function of all visible units. We connect our results to L1-regularized maximum-likelihood estimation and Chow-Liu trees. Based on our theoretical findings, we present a new tractable version of Deep Boltzmann Machines, namely the Deep Boltzmann Tree<\/em> (DBT). We provide a hyper-parameter-free algorithm for learning the DBT from data, and propose a new initialization method to enforce convergence to good solutions. Our findings provide some theoretical evidence for why a deep model might be beneficial. Experimental results on benchmark data show, that the DBT is a theoretical sound alternative to likelihood-free generative models.<\/p>","_links":{"self":[{"href":"https:\/\/randomfields.org\/index.php?rest_route=\/wp\/v2\/pages\/13"}],"collection":[{"href":"https:\/\/randomfields.org\/index.php?rest_route=\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/randomfields.org\/index.php?rest_route=\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/randomfields.org\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/randomfields.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=13"}],"version-history":[{"count":24,"href":"https:\/\/randomfields.org\/index.php?rest_route=\/wp\/v2\/pages\/13\/revisions"}],"predecessor-version":[{"id":112,"href":"https:\/\/randomfields.org\/index.php?rest_route=\/wp\/v2\/pages\/13\/revisions\/112"}],"wp:attachment":[{"href":"https:\/\/randomfields.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=13"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}