Temporal Fusion Transformer

GPTKB entity

Statements (27)
Predicate Object
gptkbp:instanceOf gptkb:model
gptkbp:arXivID 1912.09363
gptkbp:citation 2020
Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting
gptkbp:designedFor multi-horizon time series forecasting
gptkbp:developedBy gptkb:Bryan_Lim
Nicolas Loeff
Sercan Ö. Arik
Tomas Pfister
gptkbp:hasFeature quantile regression
scalable to large datasets
handling missing data
handling static and time-varying covariates
interpretable attention weights
multi-step forecasting
https://www.w3.org/2000/01/rdf-schema#label Temporal Fusion Transformer
gptkbp:implementedIn gptkb:PyTorch
gptkbp:input multivariate time series
gptkbp:openSource yes
gptkbp:output future time series predictions
gptkbp:publishedIn gptkb:NeurIPS_2020
gptkbp:uses attention mechanism
gating mechanisms
interpretable outputs
variable selection networks
gptkbp:bfsParent gptkb:PyTorch_Forecasting
gptkbp:bfsLayer 6