Traversing Time with Multi-Resolution Gaussian Process State-Space Models

Krista Elena Longi, Jakob Lindinger, Olaf Duennbier, Melih Kandemir, Arto Klami, Barbara Rakitsch

Tutkimustuotos: Artikkeli kirjassa/raportissa/konferenssijulkaisussaKonferenssiartikkeliTieteellinenvertaisarvioitu


Gaussian Process state-space models capture complex temporal dependencies in a principled manner by placing a Gaussian Process prior on the transition function. These models have a natural interpretation as discretized stochastic differential equations, but inference for long sequences with fast and slow transitions is difficult. Fast transitions need tight discretizations whereas slow transitions require backpropagating the gradients over long subtrajectories. We propose a novel Gaussian process state-space architecture composed of multiple components, each trained on a different resolution, to model effects on different timescales. The combined model allows traversing time on adaptive scales, providing efficient inference for arbitrarily long sequences with complex dynamics. We benchmark our novel method on semi-synthetic data and on an engine modeling task. In both experiments, our approach compares favorably against its state-of-the-art alternatives that operate on a single time-scale only.
OtsikkoProceedings of The 4th Annual Learning for Dynamics and Control Conference, PMLR
ToimittajatRoya Firoozi, Negar Mehr, Esen Yel, Rika Antonova, Jeannette Bohg, Mac Schwager, Mykel Kochenderfer
KustantajaJournal of Machine Learning Research
DOI - pysyväislinkit
TilaJulkaistu - 2022
OKM-julkaisutyyppiA4 Artikkeli konferenssijulkaisuussa
TapahtumaAnnual Learning for Dynamics and Control Conference - Stanford, CA, Yhdysvallat (USA)
Kesto: 23 kesäk. 202224 kesäk. 2022
Konferenssinumero: 4


NimiProceedings of Machine Learning Research (PMLR)
ISSN (elektroninen)2640-3498


  • 113 Tietojenkäsittely- ja informaatiotieteet

Siteeraa tätä