Forecasting model selection through out-of-sample rolling horizon weighted errors.

Raul Poler Josefa Mula

Published in: Expert Syst. Appl. (2011)

Keyphrases

model selection
rolling horizon
lot sizing
cross validation
single machine
hyperparameters
sample size

bayesian learning
machine learning
parameter estimation
mixture model
regression model
model selection criteria
variable selection

statistical learning
feature selection
scheduling problem
selection criterion
statistical inference
automatic model selection
motion segmentation

generalization error
support vector regression
error estimation
bayesian networks
parameter determination
leave one out cross validation