
Homework
NOTE
General
 Make sure that you have read the section about grading and the policies and guidelines.
 If you have any questions regarding a particular assignment
please contact the corresponding tutor first!
 Here you can see the current
scores
 We also maintain an archive of
previous assignments.
 There will be office hours of the tutor where the students can
ask questions regarding the homework assignments (Infohour) and
discuss the corrected solutions (Einsichtnahme). The particular
dates, times, and the room are given at the assignment.
Homework for this semester
Nmr 

Title 
Assignment 
Deadline 
Points 
Tutor 
Additional 
Solution 
1 
! 
Linear Regression
and Gradient descent 
pdf,
ps.gz 
05.05 
12.5 
Sabine
Sternig 


2 
! 
Neural
Networks 
pdf,
ps.gz 
16.05 
12.5 
Robert
Peharz 


3 
! 
Neural Networks
and Regularization 
pdf,
ps.gz 
16.05 
12.5 
Roland
Unterberger 


4 
! 
Decision Trees
and Support Vector Machines 
pdf,
ps.gz 
26.05 
12.5 
Martin
Stadler 


5 
! 
VCDimension and
Proofs 
pdf,
ps.gz 
30.05 
12.5 
Gerhard
Neumann 


6 
! 
Gaussian
Statistics and Unsupervised Learning 
pdf,
ps.gz 
06.06 
12.5 
Michael
Wiesenegger 


7 
! 
Principal
Component Analysis (PCA) and Independent Component Analysis
(ICA) 
pdf,
ps.gz 
13.06 
12.5 
Johannes
Schweighofer 


8 
! 
Hidden Markov
Models 
pdf,
ps.gz 
13.06 
12.5 
Susanne Rexeis 


9 
! 
Mixtures of
Gaussians and HMMs 
pdf,
ps.gz 
20.06 
12.5 
Susanne Rexeis 


10 
! 
Adaptive
Filters 
pdf,
ps.gz 
30.06 
12.5 
Paul Meissner 


 Since SS08 you are NOT allowed to build teams any more. Please
solve the homework individually.
 The solutions have to be handed in as hard copy.
Sometime you will be asked to hand in short Matlab programs. In
this case you have to provide a printout of the program and send
the code via email to the address given in the assignment.
 There are 10 assignments, each of them with 12.5 points. To
reach the maximum number of points is 100, thus you are only
required to hand in 8 assignments to achieve the maximum number of
points.
 The practical course is divided into 2 parts, the IGI part
(first 5 assignments) and the SPSC part. You are required to reach
at least 25 points (= 50 %) for each part in order to get a
positive grade.
 Copying solutions: If you hand in a solution
which is obviously a copy of a solution of some other student of
the course (i.g. appearance of long identical strings) the
following applies:
 zero score for this homework for each involved student (it does
not matter whose solution is the copy or the original).
 If this occurs repeatedly you will get a negative score (the
maximum score for the homework in question).
 Please write down your solutions in a
clean, structured and readable way!
 Unreadable and unintelligible/unclear solutions yield zero
score.
 Do not use any red color!
 Staple the solutions for each assigned
problem separately!
 The title page of each solution handed in has to
contain a header of the following form: HW_cover.pdf
Name 
First Name 
MatrikelNmr 




