For discussions related to modeling, machine learning and deep learning. Related packages include `caret`

, `modelr`

, `yardstick`

, `rsample`

, `parsnip`

, `tensorflow`

, `keras`

, `cloudml`

, and `tfestimators`

.

the time series model (which doesn't have periodic component) can be decomposed to

Y = T + S + R

Y = T.S.R

Y = T.S + R

Y = T + S.R (T = trend, S = Seasonal, R = Random)

But forecast package gives only additive and multiplicative models but not the combination

I want to try all 4 combinations and take the best fit. How can I do it and compare ?