Hierarchical Proxy Modeling for Improved HPO in Time Series Forecasting
Open Access
- 4 August 2023
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 891-900
Abstract
Selecting the right set of hyperparameters is crucial in time series forecasting. The classical temporal cross-validation framework for hyperparameter optimization (HPO) often leads to poor test performance because of a possible mismatch between validation and test periods. To address this test-validation mismatch, we propose a novel technique, H-Pro to drive HPO via test proxies by exploiting data hierarchies often associated with time series datasets. Since higher-level aggregated time series often show less irregularity and better predictability as compared to the lowest-level time series which can be sparse and intermittent, we optimize the hyperparameters of the lowest-level base-forecaster by leveraging the proxy forecasts for the test period generated from the forecasters at higher levels. H-Pro can be applied on any off-the-shelf machine learning model to perform HPO. We validate the efficacy of our technique with extensive empirical evaluation on five publicly available hierarchical forecasting datasets. Our approach outperforms existing state-of-the-art methods in Tourism, Wiki, and Traffic datasets, and achieves competitive result in Tourism-L dataset, without any model-specific enhancements. Moreover, our method outperforms the winning method of the M5 forecast accuracy competition.Keywords
All Related Versions
This publication has 10 references indexed in Scilit:
- Efficient Automated Deep Learning for Time Series Forecasting2023
- AutoForecast: Automatic Time-Series Forecasting Model Selection2022
- M5 accuracy competition: Results, findings, and conclusionsInternational Journal of Forecasting, 2022
- A machine learning approach for forecasting hierarchical time seriesExpert Systems with Applications, 2021
- Informer: Beyond Efficient Transformer for Long Sequence Time-Series ForecastingProceedings of the AAAI Conference on Artificial Intelligence, 2021
- DeepAR: Probabilistic forecasting with autoregressive recurrent networksInternational Journal of Forecasting, 2020
- Regularized Regression for Hierarchical Forecasting Without Unbiasedness Conditions2019
- Optimal Forecast Reconciliation for Hierarchical and Grouped Time Series Through Trace MinimizationJournal of the American Statistical Association, 2018
- The theta model: a decomposition approach to forecastingInternational Journal of Forecasting, 2000
- Conditional Heteroscedasticity in Time Series of Stock Returns: Evidence and ForecastsThe Journal of Business, 1989