Description Usage Arguments Details Value Note Author(s) References See Also
View source: R/PointProcessLasso.R
The function ppLasso
fits a generalized linear point process model
with a lasso penalization on all parameters.
1 |
formula |
an object of class |
data |
an object of class |
family |
an object of class
|
support |
a |
N |
a |
Delta |
a |
... |
additional parameters that are passed on to
|
The function provides an interface for using the glmnet
function from
the suggested package glmnet
to fit a model with the lasso penalization on the
parameter vector.
With family = Hawkes("identity")
(the default) ppLasso
uses a quadratic contrast
function for estimation, as considered in Hansen, Reynaud-Bouret and Rivoirard, and
the different parameters are weigthed according to the criteria derived in that
paper. With family = Hawkes("log")
the likelihood is used. A one-dimensional tuning
parameter remains to be selected, which is currently done by an ad hoc degrees-of-freedom
computation.
A model of class PointProcessModel
.
With family = Hawkes("identity")
the model returned may
have filter functions taking negative values. No approximate standard errors
are currently computed.
Niels Richard Hansen Niels.R.Hansen@math.ku.dk.
N. R. Hansen, P. Reynaud-Bouret and V. Rivoirard. Lasso and probabilistic inequalities for multivariate point processes. arXiv:1208.0570
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.