The course aims to teach students how to analyze single-trial high dimensional neural data. By the end of the course, students will be able to understand fundamental conceptual questions in systems neuroscience, define neural task complexity, apply low-rank matrix perturbation theory, decode neural data, recover latent states, and analyze neural dynamics using tensor components analysis. The teaching method includes lectures, simulations, and data analysis of monkey data. This course is intended for students and professionals interested in neuroscience, neural data analysis, and computational biology.
Stanford Seminar - Towards Theories of Single-Trial High Dimensional Neural Data Analysis
Stanford University via YouTube
Overview
Syllabus
Introduction.
A major conceptual elephant in systems neuroscience.
Talk outline.
Fundamental conceptual questions.
New definition of neural task complexity.
Neural Dimensionality and Task Complexity: Intuition.
A larger context: random projections.
Conclusions.
Towards a single trial theory.
Low-rank Matrix Perturbation Theory.
Static Decoding - Recovering Dimensionality.
Latent State Recovery: Simulations.
Latent State Recovery: Monkey Data.
Discovering structure in subsampled neural dynamics.
Data often modeled using latent linear dynamical systems.
Tensor components analysis.
Taught by
Stanford Online
