We will use a spontaneous activity recording from Syeda et al, 2023. We recorded 34,086 neurons from mouse sensorimotor cortex for 2+ hours using two-photon calcium imaging at a rate of 3.2Hz. In the notebook, we will build several encoding models of neural activity based on the orofacial behaviors of mice. The encoding models are increasingly more complicated:
- linear regression from spatial keypoints
- linear regression from spatiotemporal keypoints
- nonlinear regression (i.e. neural networks) from spatiotemporal keypoints
We will use rastermap to visualize the neural activity, and also to visualize the prediction.