The underlying assumptions of traditional autoregressive models are well known. The resulting complexity with these models leads to observations such as, ``We have found that choosing the wrong model or parameters can often yield poor results, and it is unlikely that even experienced analysts can choose the correct model and parameters efficiently given this array of choices.’’ Source

`NNS`

simplifies the forecasting process. Below are some examples demonstrating `NNS.ARMA`

and its assumption free, minimal parameter forecasting method.

`NNS.ARMA`

has the ability to fit a linear regression to the relevant component series, yielding very fast results. For our running example we will use the `AirPassengers`

dataset loaded in base R.

We will forecast 44 periods `h = 44`

of `AirPassengers`

using the first 100 observations `training.set = 100`

, returning estimates of the final 44 observations. We will then test this against our validation set of `tail(AirPassengers,44)`

.

Since this is monthly data, we will try a `seasonal.factor = 12`

.

Below is the linear fit and associated root mean squared error (RMSE) using `method = "lin"`

.

`nns = NNS.ARMA(AirPassengers, h = 44, training.set = 100, method = "lin", plot = TRUE, seasonal.factor = 12, seasonal.plot = FALSE)`

`sqrt(mean((nns - tail(AirPassengers, 44)) ^ 2))`

`## [1] 35.39965`

Now we can try using a nonlinear regression on the relevant component series using `method = "nonlin"`

.

`nns = NNS.ARMA(AirPassengers, h = 44, training.set = 100, method = "nonlin", plot = TRUE, seasonal.factor = 12, seasonal.plot = FALSE)`

`sqrt(mean((nns - tail(AirPassengers, 44)) ^ 2))`

`## [1] 24.34268`

We can test a series of `seasonal.factors`

and select the best one to fit. The largest period to consider would be `0.25 * length(variable)`

, in our case 25. Remember, we are testing the first 100 observations of `AirPassengers`

, not the full 144 observations.

```
seas = t(sapply(1 : 25, function(i) c(i, sqrt(mean((NNS.ARMA(AirPassengers, h = 44, training.set = 100, method = "lin", seasonal.factor = i, plot=FALSE) - tail(AirPassengers, 44)) ^ 2)))))
colnames(seas) = c("Period", "RMSE")
seas
```

```
## Period RMSE
## [1,] 1 75.67783
## [2,] 2 75.71250
## [3,] 3 75.87604
## [4,] 4 75.16563
## [5,] 5 76.07418
## [6,] 6 70.43185
## [7,] 7 77.98493
## [8,] 8 75.48997
## [9,] 9 79.16378
## [10,] 10 81.47260
## [11,] 11 106.56886
## [12,] 12 35.39965
## [13,] 13 90.98265
## [14,] 14 95.64979
## [15,] 15 82.05345
## [16,] 16 74.63052
## [17,] 17 87.54036
## [18,] 18 74.90881
## [19,] 19 96.96011
## [20,] 20 88.75015
## [21,] 21 100.21346
## [22,] 22 108.68674
## [23,] 23 85.06430
## [24,] 24 35.49018
## [25,] 25 75.16192
```

Now we know `seasonal.factor = 12`

is our best fit, we can see if there’s any benefit from using a nonlinear regression. Alternatively, we can define our best fit as the corresponding `seas$Period`

entry of the minimum value in our `seas$RMSE`

column.

You may experience instances with monthly data that report `seasonal.factor`

close to multiples of 3, 4, 6 or 12. For instance, if the reported `seasonal.factor = {37, 47, 71, 73}`

use `(seasonal.factor = c(36, 48, 72))`

. The same suggestion holds for daily data and multiples of 7, or any other time series with logically inferred cyclical patterns.

`a = seas[which.min(seas[ , 2]), 1]`

Below you will notice the use of `seasonal.factor=a`

generates the same output.

`nns = NNS.ARMA(AirPassengers, h = 44, training.set = 100, method = "nonlin",seasonal.factor = a, plot = TRUE, seasonal.plot = FALSE)`

`sqrt(mean((nns - tail(AirPassengers, 44)) ^ 2))`

`## [1] 24.34268`

There is a benefit to using a nonlinear regression as our RMSE has been lowered. We can also test if using both linear and nonlinear estimates combined result in a lower RMSE (`method = "both"`

).

`nns = NNS.ARMA(AirPassengers, h = 44, training.set = 100, method = "both", seasonal.factor = a, plot = TRUE, seasonal.plot = FALSE)`

`sqrt(mean((nns - tail(AirPassengers, 44)) ^ 2))`

`## [1] 26.31436`

Using `method = "both"`

did not lower our RMSE. There are far fewer parameters to test using NNS than traditional methods and the relative simplicity of the method ensures robustness.

`seasonal.factor`

NNS also offers a wrapper function `NNS.ARMA.optim()`

to test a given vector of `seasonal.factor`

and returns the lowest SSE and the optimal periods, as well as the `NNS.ARMA`

regression method used.

Given our monthly dataset, we will try multiple years by setting `seasonal.factor = seq(12, 48, 12)`

.

```
nns.optimal = NNS.ARMA.optim(AirPassengers, training.set = 100, seasonal.factor = seq(12, 48, 12), method = "comb")
nns.optimal
```

```
## $periods
## [1] 12
##
## $SSE
## [1] 26072.9
##
## $method
## [1] "nonlin"
```

Using our new parameters yields the same results:

`nns = NNS.ARMA(AirPassengers, training.set = 100, h = 44, seasonal.factor = nns.optimal$periods, method = nns.optimal$method, plot = TRUE, seasonal.plot = FALSE)`

`sqrt(mean((nns - tail(AirPassengers, 44)) ^ 2))`

`## [1] 24.34268`

Use of a `seasonal.factor`

too large will result in the following error. Please reduce the maximum `seasonal.factor`

```
NNS.ARMA.optim(AirPassengers,training.set = 132,seasonal.factor = seq(12, 72, 12))
Error in NNS.ARMA.optim(AirPassengers, training.set = 132, seasonal.factor = seq(12, :
Please set maximum [seasonal.factor] to less than 72
```

Using our cross-validated parameters (`seasonal.factor`

and `method`

) we can forecast another 50 periods out-of-range (`h = 50`

), by dropping the `training.set`

parameter.

`NNS.ARMA(AirPassengers, h = 50, seasonal.factor = nns.optimal$periods, method = nns.optimal$method, plot = TRUE, seasonal.plot = FALSE)`

`seasonal.factor = c(1, 2, ...)`

We included the ability to use any number of specified seasonal periods simultaneously, weighted by their strength of seasonality. Computationally expensive when used with nonlinear regressions and large numbers of relevant periods.

`seasonal.factor = FALSE`

We also included the ability to use all detected seasonal periods simultaneously, weighted by their strength of seasonality. Computationally expensive when used with nonlinear regressions and large numbers of relevant periods.

`best.periods`

This parameter restricts the number of detected seasonal periods to use, again, weighted by their strength. To be used in conjunction with `seasonal.factor = FALSE`

.

`dynamic = TRUE`

This setting generates a new seasonal period(s) using the estimated values as continuations of the variable, either with or without a `training.set`

. Also computationally expensive due to the recalculation of seasonal periods for each estimated value.

`plot`

,`seasonal.plot`

and`intervals`

These are the plotting arguments, easily enabled or disabled with `TRUE`

or `FALSE`

. `seasonal.plot = TRUE`

will not plot without `plot = TRUE`

. If a seasonal analysis is all that is desired, `NNS.seas`

is the function specifically suited for that task. `intervals`

will plot the surrounding estimated values iff `intervals = TRUE & seasonal.factor = FALSE`

.

If the user is so motivated, detailed arguments and proofs are provided within the following: