[ Main page ]

Course assignment T-61.2010

The purpose of the assigment is to get acquainted with methods introduced in the lectures. In autumn 2007 there are two alternative topics: PCA/eigenfaces and k-nearest neighbour classifier.

Common instructions

You can substitute for the course assignment by attending at least three computer sessions in November-December 2007.

Assignments are supposed to be executed individually. Write your own code and document as explained below. Discussion on high-level is very useful and accepted. Deadline is 15th January 2008.

Topics

You can choose either #1 or #2:

• Topic #1: eigenfaces
  • lecture slides chapter 4
  • additional reading material: PCA and eigenfaces
  • instructions
  • Personalised dataset that has been generated on basis of student number for every student enrolled on the course.
  • modified montage-function: myMontage.m and for Octave myOctaveMontage.m
  • some links: an article in Wikipedia, a nice project work, a tutorial to PCA (PDF)
• Topic #2: k-nearest neighbour
  • lecture slides chapter 6
  • instructions
  • Personalised dataset that has been generated on basis of student number for every student enrolled on the course.
  • a function for classification: myKnnClass.m

Some help

• common questions to news group opinnot.tik.t612010, personal ones to assistant Ville.Viitaniemi () hut.fi
• UKK - FAQ (in Finnish)
• Matlab's own documentation and tutorials, "Getting Started" etc.
• Matlab exercise sessions, see WWWTopi
• A free alternative to Matlab: Octave, cf. http://www.octave.org. Octave uses very similar commands and language as Matlab.
• Other options: this kind of projects can be implemented for example with Python or C...
• How to convert to pdf?
  • LaTeX: pdflatex or (dvips and ps2pdf)
  • Word: "print PDF"
  • OpenOffice: "export to PDF"

--> [ Pääsivulle | Main page ]