Topic: expectation-maximization

A central problem in systems neuroscience is to understand the relationship between sensory stimuli, neural activity, and decision-making behavior. Latent linear dynamical systems (LDS) models are one powerful framework for characterizing these relationships. For example, LDS models can be used to describe the hidden dynamics that govern how an animal's decision-making strategy evolves continuously over time. However, fitting LDS models to data using standard methods such as the expectation-maximization (EM) algorithm is computationally expensive and prone to local optima. We address these sho

Hippocampal studies provide a rich window into the mechanisms of memory via the paradigm of spatial navigation. To interpret neural responses, it is common practice to first build encoding models using position during locomotion as the label, and then decode the position using this model. Of particular interest is the mismatch between the decoded and the actual position, for instance, replay of non-local trajectories during rest. However, it is unclear whether a priori chosen labels explain neural data the best. By imposing spatial structure to times of immobility, we may be systematically mis