COGS 501, Fall 2009 Schedule

[Note: some of these topics (and many of these links) will change, reflecting updates, improvements, and adaptation to the interests of this year's students.]

Week
Date
Description
Readings
Assignments
1. 9/14++ Overview & organization

Mathematical background
Basic: Eero Simoncelli, "Geometric Review of Linear Algebra"
Michael Jordan, "An Introduction to Linear Algebra..."
Intermediate: Connected Curriculum Project linear algebra materials
Advanced: Desmond Fearnley-Sander "Hermann Grassman and the Creation of Linear Algebra"
Stephen Gull et al. "Imaginary Numbers are not Real"
(Note that these describe types of linear algebra significantly different from those that will be used in this course. They are offered as enrichment for any students to whom standard linear algebra is old news.)
Placement test, to be completed in class.

If you miss the first class, please fill this out and return to Laurel Sweeney at IRCS.

Instructions: there are 20 questions whose answers should be obvious if you know and remember the material, and impossible otherwise. Don't spend more than 10 minutes on it. If you don't know the answer to a question, leave it blank rather than guessing.
We don't expect you to know all this materials, or even a majority of it. The point is to give us a sense of who needs (or doesn't need) help with what.

 

         
2a 9/21 Lab session Ed Neuman's Matlab tutorials #1, #2, #3

Problem set #1
Note on r2 -- how to calculate it and what it means

2b 9/24 Lecture: Linear algebra review (Same readings as lecture 1)  
         
3a   Color vision as a 3D subspace    
3b   Lab session   Problem Set #2
Hints for Problem Set #2
         
4a   Paper Discussion Henrietta Cedergren and David Sankoff, "Variable Rules: Performance as a Statistical Reflection of Competence", Language, 50(2) pp. 333-355, 1974
Background discussion.
 
4b   Lecture: Regression  
         
5a   Lab session  

Problem set #3
HW3 hints (.m file to start you off)
(special hints for users of octave)

5b   Lecture: Basic probability and statistics review

Joint probability, conditional probability, and Bayes' Theorem
Variance and covariance

 
6a   Lab session

 

 
6b   Paper discussion Michael Heilman et al., "An Analysis of Statistical Models and Features for Reading Difficulty Predction", The Third Workshop on
Innovative Use of NLP
for Building Educational
Applications, ACL-08 (link to full proceedings)
 
7a   Lecture: Subspace Methods
Introduction to Subspace Methods  
7b   Lab session   Problem set #4
8a   Lecture: Linear classifiers Classifying multivariate data  
8b   Paper discussion

Francisco Pereira et al., "Machine learning classifiers and fMRI: a tutorial overview"

(optional additional readings)

9   Lecture: Linear shift-invariant systems

Impulse Response and Convolution

(Deconvolving the hemodynamic response...)

Problem set #5
10a   Lab session  
10b   Lecture: The discrete fourier transform

Towards the Discrete Fourier Transform
Properties of the DFT/FFT

Problem set #6
11   Lab session  
12a   Lecture FIR & IIR Filters
Poles, zeros and the z-transform
Problem set #7
12b   Paper discussion G. Kochanski et al., "Loudness predicts prominence: Fundamental frequency lends little", JASA 118(2) 1038-1054 (2005).  
13a   Lecture: fMRI analysis    
13b   Paper discussion TBA