MVD - Sampling from Correlated Multivariate Discrete Random Variables

Description

You can use the software in this package to efficiently sample from
  1. correlated multivariate binary random variables (multivariate Bernoulli),
  2. correlated multivariate Poisson random variables,
  3. correlated random variables with arbitrary marginal statistics.
Applications include modeling and generating of artificial neural data. The implementation includes sampling and parameter fitting for the Dichotomized Gaussian distribution. For some parameters, this provides an efficient alternative to the maximum-entropy distribution, the Ising model. For an instruction on how to use the code, run the demo.m script.

References

J. Macke, P. Berens, A. Ecker, A. Tolias, and M. Bethge
Generating Spike Trains with Specified Correlation-Coeffcients
Neural Computation, 2009
#spike train, #correlated poisson, #multivariate poisson, #noise correlations, #discretized gaussian
Code, PDF, BibTex

M. Bethge and P. Berens
Near-Maximum Entropy Models for Binary Neural Representations of Natural Images
Advances in Neural Information Processing Systems 20, 2008
#population coding, #natural image statistics, #maximum entropy
Code, PDF, BibTex

Files

mackeetal.zip (73.2 kB)

This file contains demo.m, a script to replicate the main figures of the paper Macke et al. (2009). It can also be downloaded from MATLAB Central File Exchange.
University of Tuebingen BCCN CIN MPI