Why should you care?
Sparse coding is both a fundamental principle for understanding sensory processing in the brain and an important machine learning technique in its own right.
What is in the tutorial?
Generally, sparse coding is a method to find the underlying factors that generate our data. There are two concrete examples in this tutorial:
First, sparse coding allows us to decompose handwritten digits into their generative components, which are pen strokes. Second, you will see how the properties of the cells in the human visual cortex are predicted by sparse coding of natural images.
Installation:
After installing python, just download the repository and open the main file 'sparse_coding.ipynb' using ipython notebook.
Uses:
Code for the RCTN section of the HWNI Neuroscience Boot Camp Tutorial for the UCB Neuroscience Data Mining Group Introduction to Sparse Coding for Graduate Students Joining the Redwood Center