« Because it's Friday: How to get a baby to sleep | Main | Exploring Global Internet Performance Data Using R »

June 20, 2016


Feed You can follow this conversation by subscribing to the comment feed for this post.

The "forecast" package is indeed the best tool out there for automatic model selection. In the context of business where many models must be automated to run on a schedule, and an expert cannot tune and select the individual models, it offers a methodologically sound approach to this problem. As an extension to this, the individual models can be combined to form a composite model that is more stable on a wide range of datasets and more accurate. This approach should be familiar to those who do machine learning and utilize ensembles/stacking. See the "forecastHybrid" package for tools to do the ensemble automatically.

The comments to this entry are closed.

Search Revolutions Blog

Got comments or suggestions for the blog editor?
Email David Smith.
Follow revodavid on Twitter Follow David on Twitter: @revodavid
Get this blog via email with Blogtrottr