Prediction Models

ForeTiS includes various time series forecasting models, both classical forecasting models as well as machine and deep learning-based methods. In the following pages, we will give some details for all of the currently implemented models. We further included a subpage explaining the Bayesian optimization that we use for our automatic hyperparameter search.

We provide both a workflow running ForeTiS with a command line interface using Docker and as a pip package, see the following tutorials for more details:

• HowTo: Run ForeTiS using Docker

• HowTo: Use ForeTiS as a pip package

In both cases, you need to select the prediction model you want to run - or also multiple ones within the same optimization run. A specific prediction model can be selected by giving the name of the .py file in which it is implemented (without the .py suffix). For instance, if you want to run Extreme Gradient Boost implemented in xgboost.py, you need to specify xgboost.

In the following table, we give the keys for all prediction models as well as links to detailed descriptions and the source code:

Time Series Forecasting Models

Model	Key in ForeTiS	Description	Source Code
Automatic Relevance Determination Regression	ard	Bayesian Regression	ard.py
SARIMA	sarima	SARIMA(X)	sarima.py
SARIMAX	sarimax	SARIMA(X)	sarimax.py
Average Historical	averagehsitorical	Baseline Models	averagehistorical.py
Average Moving	averagemoving	Baseline Models	averagemoving.py
Average Seasonal	averageseasonal	Baseline Models	averageseasonal.py
Average Seasonal Lag	averageseasonallag	Baseline Models	averageseasonallag.py
Bayesian Ridge Regression	bayesridge	Bayesian Regression	bayesridge.py
Exponential Smoothing	es	Exponential Smoothing	es.py
EVARS-GPR	evars-gpr	Gaussian Process Regression	evars-gpr.py
Gaussian Process Regression (TensorFlow Implemetation)	gprtf	Gaussian Process Regression	gprtf.py
Lasso Regression	lasso	Linear Regression	lasso.py
Long Short-Term Memory (LSTM) Network	lstm	LSTM Network	lstm.py
Bayesian Long Short-Term Memory (LSTM) Network	lstmbayes	LSTM Network	lstmbayes.py
Elastic Net Regression	elasticnet	Linear Regression	elasticnet.py
Multilayer Perceptron	mlp	Multilayer Perceptron	mlp.py
Bayesian Multilayer Perceptron	mlpbayes	Multilayer Perceptron	bayesmlp.py
Ridge Regression	ridge	Linear Regression	ridge.py
XGBoost	xgboost	XGBoost	xgboost.py

Also, ForeTiS contains the follwing additional models. These are easy to integrate by just copy and pasting them to the models folder. They do not get support or updates anymore by the authors, as we decided to use other framework for the algorithms, but they still should work.

Additional Time Series Forecasting Models

Model	Key in ForeTiS	Description	Source Code
Gaussian Process Regression (with sklearn framework)	gpr_sklearn	Gaussian Process Regression	gpr_sklearn.py
Bayesian Multilayer Perceptron (with IntelLabs BayesianTorch framework)	mlpbayes_intel	Multilayer Perceptron	mlpbayes_intel.py
Bayesian Long Short-Term Memory (LSTM) Network (with IntelLabs BayesianTorch framework)	lstmbayes_intel	LSTM Network	lstmbayes_intel.py

If you are interested in adjusting an existing model or its hyperparameters: HowTo: Adjust existing prediction models and their hyperparameters.

If you want to integrate your own prediction model: HowTo: Integrate your own prediction model.