Deep Generative Quantile-Copula Models for Probabilistic Forecasting

2019/07/24 by Wen, Ruofeng, Torkkola, Kari · 2 citations
#FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML)

paper · doi:10.48550/arxiv.1907.10697

Abstract

We introduce a new category of multivariate conditional generative models and demonstrate its performance and versatility in probabilistic time series forecasting and simulation. Specifically, the output of quantile regression networks is expanded from a set of fixed quantiles to the whole Quantile Function by a univariate mapping from a latent uniform distribution to the target distribution. Then the multivariate case is solved by learning such quantile functions for each dimension's marginal distribution, followed by estimating a conditional Copula to associate these latent uniform random variables. The quantile functions and copula, together defining the joint predictive distribution, can be parameterized by a single implicit generative Deep Neural Network.

Cited by

• Efficiently Generating Correlated Sample Paths from Multi-step Time Series Foundation Models
• From Samples to Scenarios: A New Paradigm for Probabilistic Forecasting

Related

• Probabilistic Time Series Forecasting with Implicit Quantile Networks
• Multivariate Quantile Function Forecaster
• Generative Quantile Regression with Variability Penalty
• Generative Quantile Regression with Variability Penalty
• Generative Quantile Bayesian Prediction