A self-supervised loss integrates hierarchy reconciliation into training for time series models, producing reconciled forecasts with uncertainty estimates that improve on post-hoc methods in synthetic tests.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A Self-supervised Approach to Hierarchical Forecasting with Applications to Groupwise Synthetic Controls
A self-supervised loss integrates hierarchy reconciliation into training for time series models, producing reconciled forecasts with uncertainty estimates that improve on post-hoc methods in synthetic tests.