Learning quadratic receptive fields from neural responses to natural stimuli
| Autor: | Olivier Marre, Kanaka Rajan, Gašper Tkačik |
|---|---|
| Přispěvatelé: | Princeton University, Institut de la Vision, Institut National de la Santé et de la Recherche Médicale (INSERM)-Sorbonne Université (SU)-Centre National de la Recherche Scientifique (CNRS), Centre Hospitalier National d'Ophtalmologie des Quinze-Vingts (CHNO), Institute of Science and Technology [Klosterneuburg, Austria] (IST Austria), Marre, Olivier |
| Rok vydání: | 2013 |
| Předmět: |
[SDV]Life Sciences [q-bio]
Entropy MESH: Neurons Information Theory Inference Information theory computer.software_genre MESH: Linear Models 0302 clinical medicine Quadratic equation MESH: Animals Mathematics Neurons 0303 health sciences Likelihood Functions MESH: Information Theory Covariance MESH: Entropy [SDV] Life Sciences [q-bio] MESH: Nonlinear Dynamics Neurons and Cognition (q-bio.NC) Generalized linear model MESH: Bayes Theorem Cognitive Neuroscience Bayesian probability Models Neurological Stimulus (physiology) Machine learning 03 medical and health sciences MESH: Computer Simulation Arts and Humanities (miscellaneous) MESH: Models Neurological Animals Humans Learning Computer Simulation 030304 developmental biology MESH: Humans Quantitative Biology::Neurons and Cognition business.industry Pattern recognition Bayes Theorem Nonlinear Dynamics Receptive field Quantitative Biology - Neurons and Cognition FOS: Biological sciences Linear Models MESH: Learning MESH: Likelihood Functions Artificial intelligence business computer 030217 neurology & neurosurgery |
| Zdroj: | Neural Computation Neural Computation, 2013, 25 (7), pp.1661-1692. ⟨10.1162/NECO_a_00463⟩ |
| ISSN: | 1530-888X 0899-7667 |
| DOI: | 10.1162/NECO_a_00463⟩ |
| Popis: | Models of neural responses to stimuli with complex spatiotemporal correlation structure often assume that neurons are only selective for a small number of linear projections of a potentially high-dimensional input. Here we explore recent modeling approaches where the neural response depends on the quadratic form of the input rather than on its linear projection, that is, the neuron is sensitive to the local covariance structure of the signal preceding the spike. To infer this quadratic dependence in the presence of arbitrary (e.g. naturalistic) stimulus distribution, we review several inference methods, focussing in particular on two information-theory-based approaches (maximization of stimulus energy or of noise entropy) and a likelihood-based approach (Bayesian spike-triggered covariance, extensions of generalized linear models). We analyze the formal connection between the likelihood-based and information-based approaches to show how they lead to consistent inference. We demonstrate the practical feasibility of these procedures by using model neurons responding to a flickering variance stimulus. Comment: Review, 17 pages |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|