Bayesian Learning of Loglinear Models for Neural Connectivity
This provides statistical tools for neuroscientists to detect changes in neural firing patterns with stimuli, though it is incremental as it builds on existing Boltzmann machine models.
The paper tackles the problem of learning neural connectivity structures from configuration frequency data, using a Bayesian approach that generalizes Boltzmann machines to higher-order interactions, and confirms that different attentional states in a rhesus monkey are associated with distinct interaction structures.
This paper presents a Bayesian approach to learning the connectivity structure of a group of neurons from data on configuration frequencies. A major objective of the research is to provide statistical tools for detecting changes in firing patterns with changing stimuli. Our framework is not restricted to the well-understood case of pair interactions, but generalizes the Boltzmann machine model to allow for higher order interactions. The paper applies a Markov Chain Monte Carlo Model Composition (MC3) algorithm to search over connectivity structures and uses Laplace's method to approximate posterior probabilities of structures. Performance of the methods was tested on synthetic data. The models were also applied to data obtained by Vaadia on multi-unit recordings of several neurons in the visual cortex of a rhesus monkey in two different attentional states. Results confirmed the experimenters' conjecture that different attentional states were associated with different interaction structures.