PP: GRU-ODE-Bayes: Continuous modeling of sporadically-observed time series

标签：esc   put   lin   ror   ora   dynamic   -o   ati   lua   

From: KU Leuven; ESAT-STADIUS比利时鲁汶大学

?? How to model real-world multidimensional time series? especially, when these are sporadically observed data. 

?? how to describe the evolution of the probability distribution of the data?  ODE dynamics.

sporadically-observed time series: sampling is irregular both in time and across dimensions. 

Evaluation on both synthetic data and real-world data.

Combine GRU-ODE and GRU-Bayes into GRU-ODE-Bayes model. 

Introduction: 

most methodology assumption: signals are measured systematically at fixed time intervals. 

However, most real-world data is sporadic. 

fixed time intervals data VS sporadic data.  

How to model sporadic data becomes a challenge. 

neural ordinary differential equation model; It opens the perspective of tackling the issue of irregular sampling. 

interleave the ODE and the input processing steps; + GRU + Bayesian update network. 

Performance metric: MSE, mean square error; NegLL, non-negative log-likelihood. 

?? 可是他解决了一个什么问题还不知道，只知道 是model sporadical time series. 

PP: GRU-ODE-Bayes: Continuous modeling of sporadically-observed time series

标签：esc   put   lin   ror   ora   dynamic   -o   ati   lua   

原文地址：https://www.cnblogs.com/dulun/p/12297378.html