- Lecture 1
- Lecture 2
Supplemental notes on Binary Markov Chain derivation.
- Lecture 3
- Lecture 4
Supplemental notes on Decision theory perspective on classification
- Lecture 5
Supplemental notes on Least squares derivation and the pseudoinverse; bed-time reading on the Moore-Penrose pesudoinverse (optional)
- Lecture 6
Supplemental (but vital) notes on Principal Components Analysis
- Lecture 7
Demonstration of paper presentation, on "Neurosynth" Yarkoni et al, Nature Methods Nature Methods, 8(8):665–670, 2011.
- Lecture 8
- Lab 1: Lab01.pdf
- Lab 2: Lab02.pdf
- Lab 3: Lab03.pdf
- Lab 4: Lab04.pdf Lab04_hints.pdf
Files: digits.mat display_digit.m
- Lab 5: Lab05.pdf
Files: Lab05.m GetDat.m Train.mat Test.mat
- Lab 6: Lab06.pdf Lab06_pseudocode.m
- Lab 7: Lab07.pdf
- Lab 8: Lab08.pdf Lab08_hints.pdf (including some figures of the 'solution')
Files: (see solutions zip)
- Lab 9: Lab09_CourseReview.pdf
- HW01.pdf (Roughly corresponds to Lecture 1)
- HW02.pdf (Roughly corresponds to Lecture 2)
- HW03.pdf (Roughly corresponds to Lectures 3 & 4)
- HW04.pdf (Roughly corresponds to Lectures 4 & 5)
- See Lab09_CourseReview.pdf for more review problems.
Critical reading assignment
- List of potential papers for presentation.
- Presentation Tips
- Polished, practiced and organised presentation
- Slides without too much text (short phrases, not full sentences)
- Ample but purposeful use of illustrations/graphics/figures
- Clear statement of contributions and weakness of the paper
- Enough background to make the goal/importance of the paper clear
- An introduction to matrix derivatives
- A useful matrix "cheat sheet" (from Sam Roweis, Toronto)
- Two exhaustive references: Matrix Cookbook (online book by Petersen & Pedersen) and Matrix Reference Manual (web-book by Mike Brookes).