pillow lab @ princeton

neural coding & computation group

Receptive field estimation with localized priors - matlab code

description: Computes the empirical Bayes estimate of a neuron's linear receptive field under the automatic locality determination (ALD) prior. The prior has hyperparameters that control both sparsity (zeroing coefficients outside a local region of space-time) and smoothness (zeroing Fourier coefficients outside a local region of spatio-temporal frequences). Returns MAP estimate after maximizing the marginal likelihood (or evidence) for these hyperparameters. Returns empirical Bayes ridge regression estimate as special case.

download: ald-master.zip
clone: github
more info: Description.html

relevant publication:
  • Park M & Pillow JW (2011). Receptive field inference with localized priors. PLoS Computational Biology 7(10), 1-16 [abstract | pdf | supp inf]

  • Please report any bugs to pillow@princeton.edu