APPM 4370/5370: Computational Neuroscience (Spr2020)
Lectures: MWF 11-11:50am, view online recorded videos given below3/16: Neural networks [1.1-1.5] by Hugo Larochelle
3/18: Neural networks [1.6,2.1-2.3] by Hugo Larochelle
3/20: Neural networks [2.4-2.8] by Hugo Larochelle
3/30: Neural networks [2.9-2.11] by Hugo Larochelle and Perceptron + BackProp Intro (4370youtube)
Supplemental Videos from 3blue1brown: Backprop (Deep Learning Ch 3 + 4)
4/1: Backprop in two-layer perceptron + Motivating the sigmoid (4370youtube) and read jupyter notebook for backprop
4/3: Reservoir computing + Training recurrent neural networks (4370youtube)
4/6-4/27: Meet with me in zoom office hours (or setup a zoom appt by email) to discuss ongoing project work.
Instructor: Prof. Zack Kilpatrick (zpkilpat|at|colorado|dot|edu)
Office Hours : Mon 12-1:30pm, Tues 12-1pm, or by appt on zoom
Syllabus: Complete course information. COVID-19 Addendum
Textbook (free): Neuronal dynamics: From single neurons to networks and models of cognition
by Wulfram Gerstner, Werner M. Kistler, Richard Naud, & Liam Paninski; Cambridge University Press (2014)
Supplemental Texts:
Theoretical neuroscience by Peter Dayan & Larry F. Abbott; MIT Press (2005) Chapter 7: Network Models
Spikes, decisions, and actions: Dynamical foundations of neuroscience
by Hugh R. Wilson; Oxford University Press (1999).
A short course in mathematical neuroscience by Philip Eckhoff & Philip Holmes
A probability primer by Bruno Olshausen
Prerequisites : Diff Eqns (APPM 2360 or equiv) and Matrix Methods (APPM 3310 or equiv).
Some background in probability (e.g., APPM 3570) recommended.
Grading : Your final grade is based on six HWs (60%); four non-collaborative problems (20%); and the final project (20%).
Grades are posted on canvas.
Homework : Problems are assigned via links below. Email me your homework by 11am on the due date. Late homework is not accepted. Use either Scannable or Genius Scan if you are using your phone to take photos of your HW.
python: You will learn and use python to solve homeworks and in your project. Example code/jupyter notebooks are given at this github repo.
Projects: Click this link for a list of potential project assignments. You may also meet with me to discuss something different or select a refereed research paper to read, recapitulate, and extend.
Exams? There are no exams in this course! Just a final project and four non-collaborative problems on HW2, 3, 5, & 6.
*Assignments are due by 11am on the given date. HW2, 3, 5, & 6 have non-collaborative problems, which you may NOT discuss with your classmates. Solutions are posted on canvas.
Homework 1: Due Fri 1/24
Homework 2: Due Wed 2/5
Homework 3: Due Wed 2/19
Homework 4: Due Wed 3/4
Homework 5: Due Wed 3/25
Homework 6: Due Wed 4/8