wir bieten...
Dekobild im Seitenkopf ISMLL
Courses in summer term 2010 / MSc project Artificial Intelligence/Machine Learning

Instructor: Andre Busche
Time Series Classification (Comparison) using Time Series Shapelets and/or Time Series Motifs
  • Implement a Time Series classification algorithm (paper 1) and compare it against existing results from the group
  • If time/enough students: Mine Time Series patterns (motifs, paper 2) and use them to feed a ML classifier (as proposed in paper 3: compare the results)
  • Possibly come up with simple, but effective improvements.
  1. Ye, Lexiang, Keogh, Eamonn: Time Series Shapelets: A New Primitive for Data Mining. KDD'09
  2. Lin et al.: Finding Motifs in Time Series. SIGKDD '02
  3. Buza, Krisztian, Schmidt-Thieme, Lars: Motif-based Classification of Time Series with Bayesian Networks and SVMs. GfKl 2009
Data-Mining-Cup 2010
The Data mining Cup is a data mining challenge for students to test their skills on data mining tasks. The results are compared to other either individual students, or teams.
If you win, get 2.500 Euro!
The challenge is open for submissions until May, 13th, so major effort has to be spend in the first four weeks...
Robocode is a simple, yet challenging game where tanks battle against each other. While simple tanks can be implemented in a couple of minutes, more sophisticated ones usually require much more time of thinking
The aim of the topic is to implement an own bot to let it battle against existing ones (they can be downloaded from the web). It needs to be one of your aims to become on of the TOP-50 in one of the existing leagues.
Solve the Rubik's Cube or Sudoku with a Roboter
The description should be clear from the title, eh?
The lab will provide a LEGO mindstorms roboter which should be programmed in order to able to solve either the Rubic Cube, or Sudoku (free choice) in an automated fashion. The algorithm should be "implemented" within the roboter (which has far less RAM and processing power than a usual computer ...).
Showcase the result upon your final presentation.
Time:Mon 14:00-18:00 c.t.
Assignment:KI+ML MSc
Time: by arrangement
Last Project (AI): here
Last Project (ML): here