Neural Likelihoods via Cumulative Distribution Functions
arXiv:1811.00974
Abstract
We leverage neural networks as universal approximators of monotonic functions to build a parameterization of conditional cumulative distribution functions (CDFs). By the application of automatic differentiation with respect to response variables and then to parameters of this CDF representation, we are able to build black box CDF and density estimators. A suite of families is introduced as alternative constructions for the multivariate case. At one extreme, the simplest construction is a competitive density estimator against state-of-the-art deep learning methods, although it does not provide an easily computable representation of multivariate CDFs. At the other extreme, we have a flexible construction from which multivariate CDF evaluations and marginalizations can be obtained by a simple forward pass in a deep neural net, but where the computation of the likelihood scales exponentially with dimensionality. Alternatives in between the extremes are discussed. We evaluate the different representations empirically on a variety of tasks involving tail area probabilities, tail dependence and (partial) density estimation.
arXiv comment10 pages
References in corpus (8)
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- NICE: Non-linear Independent Components Estimation
- Don't Decay the Learning Rate, Increase the Batch Size
- MADE: Masked Autoencoder for Distribution Estimation
- RNADE: The real-valued neural autoregressive density-estimator
- A Deep and Tractable Density Estimator
- Cumulative distribution networks and the derivative-sum-product algorithm
- From CDF to PDF --- A Density Estimation Method for High Dimensional Data