Abstract

We propose a nonparametric spatio-temporal stochastic differential equation (SDE) model that can learn the underlying dynamics of arbitrary continuous-time systems without prior knowledge. We augment the input space of the drift function of an SDE with a temporal component to account for spatio-temporal patterns. The experiments on a real world data set demonstrate that the spatio-temporal model is better able to fit the data than the spatial model and also reduce the forecasting error.
Original languageEnglish
Title of host publicationNIPS 2018 Spatiotemporal Workshop
Subtitle of host publication32nd Conference on Neural Information Processing Systems (NIPS 2018), Montréal, Canada
PublisherNeural Information Processing Systems Foundation
Pages1-5
Publication statusPublished - 2018
MoE publication typeD3 Professional conference proceedings
EventNIPS Spatiotemporal Workshop - Montreal, Canada
Duration: 3 Dec 20188 Dec 2018

Workshop

WorkshopNIPS Spatiotemporal Workshop
CountryCanada
CityMontreal
Period03/12/201808/12/2018

Keywords

  • stochastic differential equation
  • gaussian processes
  • spatiotemporal drift

Fingerprint Dive into the research topics of 'A Nonparametric Spatio-temporal SDE Model'. Together they form a unique fingerprint.

  • Cite this

    Yildiz, C., Heinonen, M., & Lähdesmäki, H. (2018). A Nonparametric Spatio-temporal SDE Model. In NIPS 2018 Spatiotemporal Workshop: 32nd Conference on Neural Information Processing Systems (NIPS 2018), Montréal, Canada (pp. 1-5). Neural Information Processing Systems Foundation.