Skolan för
och kommunikation
KTH / CSC / Kurser / DD2466 / indda6a-10 / Kursanalys

Kursanalys: Andra avancerad individuell kurs i datalogi, indda6a-10


 DD2465 Advanced, Individual Course in Computer Science 6.0 cr
 DD2464 Bigger Advanced, Individual Course in Computer Science 9.0 cr
 DD2463 Individual Course in Computer Science, 15 cr
 DD2462 Second Advanced, Individual Course in Computer Science 6.0 credits

Number of completed courses during fall 2009-spring 2010:

DD2464: 42 students, 3 female
Grade statistics: 31% A, 11% B, 43% C.
DD2465/6: 7 students, 0 female
Grade statistics: 14% A, 29% B, 57% C.
DD2463: 4 students, 0 female
Grade statistics: 75% A, 25% B.

Course given: Continously

This is a series of courses with undefined topics; they should involve an 
investigation of a computer science related topic done individually or in 
a small group

Course Goals:
These must be rather generic, and is the same for all courses:

After passing this course you will be able to solve a design,
implementation or investigative task with limited supervision, and to
report the result of such a task. More precisely, you will be able to

    * plan, carry through and report such a task in Computer Science,
    * obtain and evaluate information applicable for carrying out the task,
    * choose a course of approach and define, follow and follow up a plan 
      for carrying out the task in a given resource budget (6 hp),
    * report your results orally and in writing, professionally,
    * show increased knowledge in an area of Computer Science.

Examiner: Stefan Arnborg, advisor from CSC School

Lectures: Only individual advising

Registered students: 

Course Literature:
Selected for each course participant


Oral or written project report or demo


The latest year (2009/10) the following projects (among others)  have 
been completed (some by a team of several students):

Rättning av lexikala fel i svensk text
Correction of lexical errors in Swedish text
Rendering och editering av voxelmodeller.
Consensus Propagation, Belief Propagation in Gaussian Markov Fields
Running Gaussian Process Latent Variable Model on a GPU.
Implementering och test av GPLVM på grafikprocessor.
Visualisering av Vintergatans centrum
Visualisation of the centre of the Galaxy.
Practical optimization of airline operations.
Praktisk optimering av flygtrafik.
Airplane flying game with OpenGL
Parallell programmed implementation of a mobile phone game

Some earlier projects:

Generic blogg system (2 persons)
Simulation of dynamic object manipulation in robotics
Interpretation of written symbols (OCR) on pressure-sensitive 
  mobile phone screen
Implementation of Spotify system on iPhone (Spotify, 2 pers)
Improving NetFlix recommender method (2 persons, Spotify)
Multiple alignment of DNA with confidence estimate (2 persons)
Using Bloom filter in DNA sequencing
Dynamic Cash flow supervision with web interface
Encryption of files in mobile phone application
Case description of large scale systems engineering projects
Tag the world in Google Android  application.
Manpower scheduling in 'small' company
Generation of synthetic music (comp part) by genre.
Generic object image retrieval in vision
Improved segmentation by patch clustering (2 projects)

It is difficult to ensure that student administers his/her time
adequately. Examiner relies on advisor to follow up on student.


Through this course students with a special interest area within the
field of computer science can perform studies that have been
individually defined for the specific student to fit his/her
interests. Course contents and examination will be individually
defined for each student. Students interested in taking the course are
asked to contact first the person in charge of the corresponding
specialization or some other teacher and then the person in charge of
this course. The course can only be offered if the department has
sufficient resources and competence within the special interest area.

Changes made in 2010:

Each project is unique, but the workflow has not changed in recent years.


The course appears to lie right wrt the background of this years
(2009/10) students.


The new grading system is applied so that in the majority of cases grades
3, 4 and 5 correspond to C, B, and A.  Grading criteria are mainly based on
the visible learning and presentation aspects of the project.
Project well done and with a significant research content gives grade A,
which is quite common because of the electivity of the course.


Is not used because of the nature of the course. Individual feedback
during/after examination.


Many students seem to spend more time on these course than its credit
translates to, but this is difficult to prevent wirhout too resource-consuming


The courses have only been chosen by students with adequate experience
and ability, but of course projects are less predictable than ordinary 





Copyright © Sidansvarig: Stefan Arnborg <>
Uppdaterad 2010-09-09