All Courses
-
9.S52/9.S912 Principles of Neural Computation in Brains and Machines
Brains and many machines rely on computation in neural networks. This course introduces and discusses some of the core principles behind neural computation in brains, machines, or both. Here by principle, we refer loosely to big ideas that we choose to keep if we must lose all other knowledge. The course has two components. Each component consists of a number of modules. The first component is Views, which describes different ways we can think about neural computation. This component consists of Representation, Algorithm, and Dynamics modules. The second component is Principles, which covers some basic principles of neural computation. Each module is structured in the following way: What is this principle? What is an example of this principle applied? Why is this principle fundamentally important? Intuition and/or Theory What are some other examples of this principle in brains and machines? (Optional) What is currently missing from our understanding? There will be lots of open discussions during the lectures.