我有大约10000个时间序列。
我想使用auto.arima函数http://www.inside-r.org/packages/cran/forecast/docs/auto.arima
我想测试10000个时间序列的auto.arima模型的准确性。我拿走了20％的数据点（如果你看到40个中的样本我会推迟8）然后让auto.arima预测。 然后我可以将生成的8个值与实际的8个值进行比较 但是有没有正式的方法来测试ARIMA模型的准确性？我的方法是否正确？
y=auto.arima(x)
plot(forecast(y,h=8))
样品 时间序列1
0.0003748,0.0003929,0.0003653,0.0003557,0.0004463,0.000349,0.0003099,0.0003395,0.0003157,0.0002871,0.0002604,0.0002422,0.0001917,0.0002117,0.0002689
时间序列2
0.0003977,0.0003481,0.0002413,0.0002069,0.0002127,0.0002108,0.0002003,0.0002174,0.0002098,0.0002069,0.0001955,0.0001926,0.0002108,0.0002146,0.0002079
答案 0 :(得分：0)
在我看来，您的Q就是用于比较预测准确度的不同指标，而不是auto.arima()和forecast()的具体用途。如果是这样，那么可以使用许多指标。有关概述，请参阅
https://en.wikipedia.org/wiki/Forecasting#Forecasting_accuracy
他们每个人都有其支持者和批评者;例如，见本文：
http://robjhyndman.com/papers/mase.pdf
无论您使用何种准确度指标，您仍需要能够证明为什么要阻止20％的数据进行预测。
但是，如果您对不同的模型表单感兴趣，那么您也有一些选择。例如，正如评论中所建议的那样，
arima()（或某些等价物）为每个时间序列拟合相同的单变量模型（指定先验）; auto.arima()为每个时间序列拟合（可能）不同的单变量模型;或如果您感兴趣的是＃3，我建议MARSS pkg：
https://cran.r-project.org/web/packages/MARSS/index.html
和用户指南：
https://cran.r-project.org/web/packages/MARSS/vignettes/UserGuide.pdf