Approximate leave-future-out cross-validation for Bayesian time series models

被引:54
|
作者
Burkner, Paul-Christian [1 ]
Gabry, Jonah [2 ,3 ]
Vehtari, Aki [1 ]
机构
[1] Aalto Univ, Dept Comp Sci, Konemiehentie 2, Espoo 02150, Finland
[2] Columbia Univ, Appl Stat Ctr, New York, NY USA
[3] Columbia Univ, ISERP, New York, NY USA
基金
芬兰科学院;
关键词
Time series analysis; cross-Validation; Bayesian inference; pareto Smoothed importance sampling; R PACKAGE;
D O I
10.1080/00949655.2020.1783262
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
One of the common goals of time series analysis is to use the observed series to inform predictions for future observations. In the absence of any actual new data to predict, cross-validation can be used to estimate a model's future predictive accuracy, for instance, for the purpose of model comparison or selection. Exact cross-validation for Bayesian models is often computationally expensive, but approximate cross-validation methods have been developed, most notably methods for leave-one-out cross-validation (LOO-CV). If the actual prediction task is to predict the future given the past, LOO-CV provides an overly optimistic estimate because the information from future observations is available to influence predictions of the past. To properly account for the time series structure, we can use leave-future-out cross-validation (LFO-CV). Like exact LOO-CV, exact LFO-CV requires refitting the model many times to different subsets of the data. Using Pareto smoothed importance sampling, we propose a method for approximating exact LFO-CV that drastically reduces the computational costs while also providing informative diagnostics about the quality of the approximation.
引用
收藏
页码:2499 / 2523
页数:25
相关论文
共 50 条
  • [1] Using Bayesian Leave-One-Out and Leave-Future-Out Cross-Validation to Evaluate the Performance of Rate-Time Models to Forecast Production of Tight-Oil Wells
    Maraggi L.M.R.
    Lake L.W.
    Walsh M.P.
    SPE Reservoir Evaluation and Engineering, 2022, 25 (04): : 730 - 750
  • [2] Using Bayesian Leave- One- Out and Leave-Future-Out Cross-Validation to Evaluate the Performance of Rate- Time Models to Forecast Production of Tight-Oil Wells
    Maraggi, Leopoldo M. Ruiz
    Lake, Larry W.
    Walsh, Mark P.
    SPE RESERVOIR EVALUATION & ENGINEERING, 2022, 25 (04) : 730 - 750
  • [3] Automatic cross-validation in structured models: Is it time to leave out leave-one-out?
    Adin, Aritz
    Krainski, Elias Teixeira
    Lenzi, Amanda
    Liu, Zhedong
    Martinez-Minaya, Joaquin
    Rue, Havard
    SPATIAL STATISTICS, 2024, 62
  • [4] Bayesian model averaging for mortality forecasting using leave-future-out validation
    Barigou, Karim
    Goffard, Pierre-Olivier
    Loisel, Stephane
    Salhi, Yahia
    INTERNATIONAL JOURNAL OF FORECASTING, 2023, 39 (02) : 674 - 690
  • [5] Robust Leave-One-Out Cross-Validation for High-Dimensional Bayesian Models
    Silva, Luca Alessandro
    Zanella, Giacomo
    JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 2024, 119 (547) : 2369 - 2381
  • [6] Bayesian Leave-One-Out Cross-Validation for Large Data
    Magnusson, Mans
    Andersen, Michael Riis
    Jonasson, Johan
    Vehtari, Aki
    INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 97, 2019, 97
  • [7] Limitations of Bayesian Leave-One-Out Cross-Validation for Model Selection
    Gronau Q.F.
    Wagenmakers E.-J.
    Computational Brain & Behavior, 2019, 2 (1) : 1 - 11
  • [8] Rejoinder: More Limitations of Bayesian Leave-One-Out Cross-Validation
    Gronau Q.F.
    Wagenmakers E.-J.
    Computational Brain & Behavior, 2019, 2 (1) : 35 - 47
  • [9] On cross-validation of Bayesian models
    Alqallaf, F
    Gustafson, P
    CANADIAN JOURNAL OF STATISTICS-REVUE CANADIENNE DE STATISTIQUE, 2001, 29 (02): : 333 - 340
  • [10] Efficient approximate leave-one-out cross-validation for kernel logistic regression
    Gavin C. Cawley
    Nicola L. C. Talbot
    Machine Learning, 2008, 71 : 243 - 264