A Deep and Tractable Density Estimator

نویسندگان

  • Benigno Uria
  • Iain Murray
  • Hugo Larochelle
چکیده

The Neural Autoregressive Distribution Estimator (NADE) and its real-valued version RNADE are competitive density models of multidimensional data across a variety of domains. These models use a fixed, arbitrary ordering of the data dimensions. One can easily condition on variables at the beginning of the ordering, and marginalize out variables at the end of the ordering, however other inference tasks require approximate inference. In this work we introduce an efficient procedure to simultaneously train a NADE model for each possible ordering of the variables, by sharing parameters across all these models. We can thus use the most convenient model for each inference task at hand, and ensembles of such models with different orderings are immediately available. Moreover, unlike the original NADE, our training procedure scales to deep models. Empirically, ensembles of Deep NADE models obtain state of the art density estimation performance.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

A Berry-Esseen Type Bound for a Smoothed Version of Grenander Estimator

In various statistical model, such as density estimation and estimation of regression curves or hazard rates, monotonicity constraints can arise naturally. A frequently encountered problem in nonparametric statistics is to estimate a monotone density function f on a compact interval. A known estimator for density function of f under the restriction that f is decreasing, is Grenander estimator, ...

متن کامل

Asymptotic Behaviors of Nearest Neighbor Kernel Density Estimator in Left-truncated Data

Kernel density estimators are the basic tools for density estimation in non-parametric statistics.  The k-nearest neighbor kernel estimators represent a special form of kernel density estimators, in  which  the  bandwidth  is varied depending on the location of the sample points. In this paper‎, we  initially introduce the k-nearest neighbor kernel density estimator in the random left-truncatio...

متن کامل

Neural Autoregressive Distribution Estimation

We present Neural Autoregressive Distribution Estimation (NADE) models, which are neural network architectures applied to the problem of unsupervised distribution and density esitmation. They leverage the probability product rule and a weight sharing scheme inspired from restricted Boltzmann machines, to yield an estimator that is both tractable and has good generalization performance. We discu...

متن کامل

Cascading Denoising Auto-Encoder as a Deep Directed Generative Model

Recent work (Bengio et al., 2013) has shown how Denoising Auto-Encoders(DAE) become generative models as a density estimator. However, in practice, the framework suffers from a mixing problem in the MCMC sampling process and no direct method to estimate the test loglikelihood. We consider a directed model with an stochastic identity mapping (simple corruption process) as an inference model and ...

متن کامل

Comparison of the Gamma kernel and the orthogonal series methods of density estimation

The standard kernel density estimator suffers from a boundary bias issue for probability density function of distributions on the positive real line. The Gamma kernel estimators and orthogonal series estimators are two alternatives which are free of boundary bias. In this paper, a simulation study is conducted to compare small-sample performance of the Gamma kernel estimators and the orthog...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2014