How does one add exogenous data to exponential smoothing models?

Question

Adding exogenous data as external regressors in an ARIMA model seems pretty straightforward, but how does one do so with ETS methods?

Is there a way of creating a weighted combination of the ETS forecast and a regression based on the exogenous data?

$\begingroup$ See robjhyndman.com/hyndsight/ets-regressors $\endgroup$
– Rob Hyndman
Commented Dec 13, 2016 at 10:27 — Rob Hyndman, Commented Dec 13, 2016 at 10:27

user176030 · Accepted Answer · 2017-09-04 23:13:37Z

1

Some empirical studies have been done at the following papers, to create ETS models with regressors.

answered Sep 4, 2017 at 23:13

user176030

111 bronze badge

2

$\begingroup$ Please give some details, along with complete references (if your links die your answer will be almost useless). As far as possible answers should be self-contained, using links references as support for what you say in an answer rather than as an answer. See the section under "Provide context for links" in the help on writing answers. You might also find the example of quoting material from a source (and the subsequent discussion) here helpful. $\endgroup$
– Glen_b
Commented Sep 4, 2017 at 23:59

Add a comment |

1 Answer 1