selectHyperParms | R Documentation |
Automatic selection of model hyper-parameters
selectLSTAR(x, m, d=1, steps=d, mL = 1:m, mH = 1:m, thDelay=0:(m-1),
fast=TRUE, trace=FALSE)
selectNNET(x, m, d=1, steps=d, size=1:(m+1), maxit=1e3, trace=FALSE)
x |
time series |
m , d , steps |
embedding parameters. For their meanings, see help about |
mL , mH |
Vector of ‘low’ and ‘high’ regimes autoregressive orders |
thDelay |
Vector of ‘threshold delay’ values |
size |
Vector of numbers of hidden units in the nnet model |
maxit |
Max. number of iterations for each model estimation |
fast |
For LSTAR selection, whether a fast algorithm using starting values fro previous models should be used |
trace |
Logical. Whether informations from each model should be returned. |
Functions for automatic selection of LSTAR and NNET models hyper parameters. An exhaustive search over all possible combinations of values of specified hyper-parameters is performed.
Embedding parameters m,d,steps
are kept fixed.
Selection criterion is the usual AIC.
For the LSTAR model, two methods are offered:
Each model is run separately, each time using the full grid search for starting values.
Only the first model is run with a full grid search, while the subsequent use the first model results for their starting values.
A data-frame, with columns giving hyper-parameter values and the computed AIC for each row (only the best 10s are returned)
Antonio, Fabio Di Narzo
llynx <- log10(lynx)
selectLSTAR(llynx, m=2)
selectNNET(llynx, m=3, size=1:5)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.