Autoregressive models are among the best performing neural density
estimators. We describe an approach for increasing the flexibility of an
autoregressive model, based on modelling the random numbers that the model uses
internally when generating data. By constructing a stack of autoregressive
models, each modelling the random numbers of the next model in the stack, we
obtain a type of normalizing flow suitable for density estimation, which we
call Masked Autoregressive Flow. This type of flow is closely related to
Inverse Autoregressive Flow and is a generalization of Real NVP. Masked
Autoregressive Flow achieves state-of-the-art performance in a range of
general-purpose density estimation tasks.
Beschreibung
[1705.07057] Masked Autoregressive Flow for Density Estimation
%0 Journal Article
%1 papamakarios2017masked
%A Papamakarios, George
%A Pavlakou, Theo
%A Murray, Iain
%D 2017
%K flows generative-models
%T Masked Autoregressive Flow for Density Estimation
%U http://arxiv.org/abs/1705.07057
%X Autoregressive models are among the best performing neural density
estimators. We describe an approach for increasing the flexibility of an
autoregressive model, based on modelling the random numbers that the model uses
internally when generating data. By constructing a stack of autoregressive
models, each modelling the random numbers of the next model in the stack, we
obtain a type of normalizing flow suitable for density estimation, which we
call Masked Autoregressive Flow. This type of flow is closely related to
Inverse Autoregressive Flow and is a generalization of Real NVP. Masked
Autoregressive Flow achieves state-of-the-art performance in a range of
general-purpose density estimation tasks.
@article{papamakarios2017masked,
abstract = {Autoregressive models are among the best performing neural density
estimators. We describe an approach for increasing the flexibility of an
autoregressive model, based on modelling the random numbers that the model uses
internally when generating data. By constructing a stack of autoregressive
models, each modelling the random numbers of the next model in the stack, we
obtain a type of normalizing flow suitable for density estimation, which we
call Masked Autoregressive Flow. This type of flow is closely related to
Inverse Autoregressive Flow and is a generalization of Real NVP. Masked
Autoregressive Flow achieves state-of-the-art performance in a range of
general-purpose density estimation tasks.},
added-at = {2019-12-09T11:33:05.000+0100},
author = {Papamakarios, George and Pavlakou, Theo and Murray, Iain},
biburl = {https://www.bibsonomy.org/bibtex/24ae7ea0be7214ba4452bf56fc4b3e53d/kirk86},
description = {[1705.07057] Masked Autoregressive Flow for Density Estimation},
interhash = {76f7aa89219237bf14bf0f35d8eb5649},
intrahash = {4ae7ea0be7214ba4452bf56fc4b3e53d},
keywords = {flows generative-models},
note = {cite arxiv:1705.07057Comment: section 4.3 is corrected since the previous version},
timestamp = {2019-12-09T11:33:05.000+0100},
title = {Masked Autoregressive Flow for Density Estimation},
url = {http://arxiv.org/abs/1705.07057},
year = 2017
}