Linear Modeling of Neural Response

This unit focuses on linear models for modeling the relationship between continuous neural recordings and events like stimuli.

The first two lessons will focus on the theoretical background: First you’ll learn how to apply linear regression and interpret the estimated coefficients. You’ll also see how ridge regularization can help to prevent overfitting and allow the model to make better predictions on new unseen data. The second lesson will then extend the linear model to relationships that are not instantaneous (like the neural response to stimuli) by introducing time-lags.

The third lesson will introduce mTRFpy, a package that provides a convenient interface for applying linear models to neural data and explore its application for modeling neural responses to continuous naturalistic speech.