Computational-Neuroscience-UW

The University of Washington offers an online course on computational neuroscience taught by Adrienne Fairhall and Rajesh Rao. I found out that I could gain a lot of insight by simulating the methods and models discussed. Also, having a script in which you can quickly adjust some parameters and investigate the effects really makes the material come alive :). I have made these scripts available so that other learners can also benefit!

The following content has been included so far:

• Week 2: Neurons as linear filters (simple cells, ganglion cells), how to estimate linear receptive fields using spike-triggered averaging and a demonstration of the linear-nonlinear-poisson model. Finally, based on a learner's question on the discussion board I included an example of how to compute eigenfaces!

• Week 3: Signal detection theory and maximum likelihood estimation. Contains a simulation of how the direction in which a stimulus is moving can be decoded from neural responses!

• Week 4: An introduction to information theory

• Week 5: RC-circuit models, integrate-and-fire models and the world famous Hodgkin-Huxley model of action potential generation!

• Week 6: In progress

• Week 7: In progress

• Week 8: In progress

Each folder contains python files with simulation code. These scripts are basically my 'lecture notes' and might still contain small mistakes or unclarities in the code. Starting from these lecture notes, I write IPython notebooks that should be much clearer to follow. This is all work in progress so make sure to check back regularly!

Enjoy!