SSpace: A Toolbox for State Space Modeling

Abstract

SSpace is a MATLAB toolbox for state space modeling. State space modeling is in itself a powerful and flexible framework for dynamic system modeling, and SSpace is conceived in a way that tries to maximize this flexibility. One of the most salient features is that users implement their models by coding a MATLAB function. In this way, users have complete flexibility when specifying the systems, have absolute control on parameterizations, constraints among parameters, etc. Besides, the toolbox allows for some ways to implement either non-standard models or standard models with non-standard extensions, like heteroskedasticity, time-varying parameters, arbitrary nonlinear relations with inputs, transfer functions without the need of using explicitly the state space form, etc. The toolbox may be used on the basis of scratch state space systems, but is supplied with a number of templates for standard widespread models. A full help system and documentation are provided. The way the toolbox is built allows for extensions in many ways. In order to fuel such extensions and discussions an online forum has been launched.

Supply chain decision support systems based on a novel hierarchical forecasting approach

“…A simulation of the above models produces forecasts like the ones shown in Figure 2. The standard procedure is done in the R environment with the hts package (Hyndman et al, 2011), while the SS forecasts are done in MATLAB with the SSpace toolbox (Villegas and Pedregal, 2018). The example is done in a way such that the base AR models are exactly the same in both cases, making sure that the observed differences are strictly and solely due to the aggregation procedures.…”

“…Further sophistication may be introduced in both equations by means of additional matrices taking into account either linear or non-linear input-output relationships. These particular configurations with a number of extensions not quoted here (such as non-gaussian and non-linear models) are implemented in the SSpace toolbox written by the authors for the MATLAB environment (The MathWorks, Inc, 2018; Villegas and Pedregal, 2018). This toolbox is based on several references, like Harvey (1989), Young et al (1999), Durbin and Koopman (2012), Casals et al (2016).…”

“…Datasets of 100 samples were generated, taking the first 90 samples as the training partition for identifying and fitting the models, and the last 10 observations as the out-of-sample partition. Automatic identification of ARIMA models was done using SSpace (Villegas and Pedregal, 2018) following the algorithm implemented in the forecast package in R (Hyndman and Khandakar, 2008). The fitted models were used to produce 1-to-7 steps-ahead predictions for each of the 15 series.…”

“…hts (Hyndman et al, 2011) and SSpace (Villegas and Pedregal, 2018) implementations. The simulation was repeated 1000 times.…”

On the Automatic Identification of Unobserved Components Models

“…Finally, software packages are scarcer than packages for other more standard techniques. Some complete alternatives are, for example, STAMP [18], SSfpack [19], and SSpace [26].…”