With large quantities of data typically available nowadays, forecasting models that are trained across sets of time series, known as Global Forecasting Models (GFM), are regularly outperforming traditional univariate forecasting models that work on isolated series. As GFMs usually share the same set of parameters across all time series, they often have the problem of not being localised enough to a particular series, especially in situations where datasets are heterogeneous. We study how ensembling techniques can be used with generic GFMs and univariate models to solve this issue. Our work systematises and compares relevant current approaches, namely clustering series and training separate submodels per cluster, the so-called ensemble of specialists approach, and building heterogeneous ensembles of global and local models. We fill some gaps in the approaches and generalise them to different underlying GFM model types. We then propose a new methodology of clustered ensembles where we train multiple GFMs on different clusters of series, obtained by changing the number of clusters and cluster seeds. Using Feed-forward Neural Networks, Recurrent Neural Networks, and Pooled Regression models as the underlying GFMs, in our evaluation on six publicly available datasets, the proposed models are able to achieve significantly higher accuracy than baseline GFM models and univariate forecasting methods.

中文翻译：

---

时间序列预测的本地化模型集成

如今，由于通常具有大量数据，跨时间序列集训练的预测模型（称为全球预测模型（GFM））通常胜过可用于孤立序列的传统单变量预测模型。由于GFM通常在所有时间序列中都共享相同的参数集，因此它们通常存在无法充分定位到特定序列的问题，尤其是在数据集异构的情况下。我们研究如何将集合技术与通用GFM和单变量模型一起使用以解决此问题。我们的工作对当前的相关方法进行了系统化和比较，即聚类系列和每个聚类训练单独的子模型，所谓的专家方法集成，以及构建全局和局部模型的异类集成。我们填补了这些方法中的一些空白，并将它们概括为不同的基础GFM模型类型。然后，我们提出了一种新的聚类集成方法，其中我们通过更改聚类和聚类种子的数量，在不同序列的聚类上训练多个GFM。使用前馈神经网络，递归神经网络和池回归模型作为基础GFM，在我们对六个公开可用的数据集进行评估时，所提出的模型能够比基线GFM模型和单变量预测方法实现更高的准确性。