Approximate leave-future-out cross-validation for Bayesian time series models
Approximate leave-future-out cross-validation for Bayesian time series models
复制标题
DOI：
10.1080/00949655.2020.1783262
复制
发表时间：
2020-06-25
影响因子：
1.2
通讯作者：
Vehtari, Aki
中科院分区：
文献类型：
--
作者：
Burkner, Paul-Christian;Gabry, Jonah;Vehtari, Aki
时间序列分析的一个共同目标是使用观察到的序列为未来的观测提供预测信息。在没有任何实际的新数据进行预测的情况下，交叉验证可以用于估计模型的未来预测准确性，例如，用于模型比较或选择的目的。贝叶斯模型的精确交叉验证通常在计算上是昂贵的，但是近似交叉验证方法已经被开发出来，最值得注意的是留一交叉验证（LOO-CV）方法。如果实际的预测任务是在给定过去的情况下预测未来，则LOO-CV提供了过于乐观的估计，因为来自未来观测的信息可以影响对过去的预测。为了正确地解释时间序列结构，我们可以使用留后交叉验证（leave-future-out cross-validation，LFO-CV）。与精确LOO-CV一样，精确LFO-CV需要多次根据不同的数据子集重新调整模型。使用帕累托平滑的重要性采样，我们提出了一种方法近似精确的LFO-CV，大大降低了计算成本，同时还提供了有关近似质量的信息诊断。
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.