Application of Bagged Copula-GP: Confirming Neural Dependency on Pupil Dilation
Abstract: Advancements in recording techniques has enabled the ability to record thousands of neurons simultaneously, shifting the needs within the field of computational neuroscience to powerful computational and statistical techniques. Copula-GP is a recently developed state-of-the-art parametric mutual information estimator found to outperform other novel non-parametric methods when utilized on highly dimensional data. Here, we utilized Copula-GP together with Gaussian Process Factor Analysis (GPFA) to investigate the information interaction between neuronal processes within the visual cortex of live mice and pupil dilation. We found usage of GPFA as a preprocessing step to Copula-GP was an effective means of investigating neuronal dependence, allowing flexibility in analysis and finding results in agreement with prior literature, and additionally extended Copula-GP with a bagging framework, allowing for the aggregation of model estimations and allowing for more accurate estimation accuracy and representation of dependency shape. We validated our bagging algorithm on simulated data sampled from known distributions, and utilized bagged Copula-GP with GPFA on said neuronal data to find results in agreement with baseline Copula-GP but with more stability.
Submission Length: Regular submission (no more than 12 pages of main content)
Changes Since Last Submission: - Changed pseudo-code to be more generally written.
- Fixed misplaced subsection in appendix.
- Removed a $\LaTeX$ command from the OpenrReview abstract.
Assigned Action Editor: ~Bertrand_Thirion1
Submission Number: 3216
Loading