This package contains two examples (in Matlab) of how to compute the log likelihood using Gaussian quadrature. The first example (RenewalExample.m) corresponds to a Renewal Process (with inverse Gaussian ISI-distribution). Using a simulated Renewal Process we give the approximation of the true log likelihood using the four approximations used in the article (DR1, DR2, CT and GL) The second example (ExampleGLM.m) is a simple refractory GLM model with 1-dimensional covariate and 2 parameters to be estimated, gain and log baseline firing rate. First, a spike train is simulated, and then estimation is done using Newton Method with backtracking linesearches, both for DR1 and GL approximations. For the GL method, there are several ways of computing the log likelihood evaluation nodes and weights. For example, the nodes can be obtained by finding the roots of the derivatives of the Legendre polynomials, and the weights through evaluations of some function of the polynomials and their derivatives at the corresponding roots. See lglnodes.m for computational details.
Code for the paper On quadrature methods for refractory point process likelihoods
License
gomena/RefractoryLikelihoods
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
Code for the paper On quadrature methods for refractory point process likelihoods
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published