T-106.290 Ohjelmoinnin laboratoriotyöt

Course brochure spring 2004

0. Introduction

In this course students implement a small technology evaluation or an experimental search project. In the same time students are introduced to scientific research.

Course is meant only for Computer Science students.

1. Objectives

• The course familiarize students with experimental algorithm research and methodics and problems of technology evaluations. Familiarized algorithms are mainly subject in course T-106.250 Datastructures and algorithms. It is also possible to get know other subjects if own subject is proposed.
• In the course students practice measuring, writing scientific articles, qualitative analysis and statistical processing, graphical representation and source of error analysis of measurement results. Measuring system is programmed during the course.
• While writing the article students also get to know about scientific writing, evaluating, reviewing and taking a presentation.

2 Prerequisites

• T-93.210 Ohjelmoinnin peruskurssi T1 (or the earlier version).
• T-106.230 Ohjelmoinnin peruskurssi T2 (or the earlier version).
• T-106.250 Tietorakenteet ja algoritmit (or the earlier version).

It is useful to have some basic knowledge of mathematical statistics but it can be studied during the course.

3. Teaching methods

1. Lectures
   • At the moment it seems that the course will have five lectures, that will be hold on tuesdays at lecture hall T1, time is 8-10. First lecture is 20.1. Complete lecture schedule will be published later.
   • Lectures will be mainly about:
     • Introduction to experimental algorithm research and technology evalutions. Why they are needed, what kind of methods are used and what kind of problems are found?
     • Examples of experiment setups, algorithms and results.
     • Critical examination of results.
     • Writing scientific article.
     • Free software for scientific writing: Latex, Gnuplot.
     • At least one special lecture that is about experimental research.
2. Study
   • In the study students familiarize oneself with one algorithm or computer engineering sector. This contains studying the algorithms or software, comparing them in theory (analytical results) and in practice.This means that students will implement a bunch of algorithms and perform experimental comparision with them.

4. Grading

Course is passed by doing a study that will be graded zero to five (0-5). To get pass man should have at least grade one (1). There are no exams in this course. Grade is based on theory part (weight 15%), experiment plan (15%), final report (60%) and peer review made about other groups (10%). Notice that theory part and experiment plan are 30% of the course grade.

5. Study

5.1 Objectives

Objective of the study is to teach methodics of experimental algorithm research, problems and methodics of technology evaluation. Students are also teached how process experimental data from research. Study has also other teaching aspects. Students are practicing information gathering, critical reviewing, programming, scientific writing and presentation of results to audience.

5.2 Implementation

• Study can be 2-5 credits depending on the subject.
• Course has list of subjects. It is also possible and recommended to make a study from own subject if some special algorithm area is in mind. If students want to make a study about own subject, first read the other subjects and then make a similar subject succestion that should be send to course lecturer for comments and approval.
• Study is made in groups of 2-3 persons..
• Studies are guided by course staff.
• Every group will make a www-page for the course documents. More instructions about how to return documents can be found on the course homepage. Exact deadlines are also on the course homepage.
• All study reports are written with Latex word processing system, because Latex is essential tool in scientific writing.
• Literary research and experimental plan are made in the beginning of the study.
• • In literary research students familiriaze themselves with the algorithms and/or software that are given in the topic. Sourcecode or pseudocode of the studied algorithms are found and software that will be studied will be installed.
  • In experimental plan students show specific plans and objectives of how the studied variables effect the test results. Objectives have to be justfied carefully.
• A meeting is hold with the course staff when doing the experimental plan.
• Course assistants will grade theory part and experimental plan and when needed will call the group for a meeting where possible problems of the plans are discussed.
• Study is written in final report, that contains:
• • Motivation and introduction to the problem studied
  • Experiment results in graphs.
  • Analysis of possible source of errors.
  • Conclusions of the results.
• After returning the final report students review reports of the other groups and give written feedback. Students can improve their report based on the given feedback. After that the final version is evaluated by course staff.
• Organizers of the course can possibly make a summary publication of the best reports.
• The best reports are presented to other students or interested parties in a mini conference. Abstracts and reports of the best studies are also showed in the course homepage.

6. Study material and communications

There is no course book. Some handouts will be delivered.

7. Course staff

Leader of the course is researcher Kenneth Oksanen, who is met only by agreement. Oksanen and assistants Ville Nenonen, Mikko Rahikainen, Jussi Rautio are guiding through the study and are also grading the course work. Every group will have their own responsible assistant that will work with the group during the course.

Course news are announced at the lectures and in the front page of course homepage. Courses homepage address is http://www.cs.hut.fi/Opinnot/T-106.290/.

The course personnel does not use any news groups for questions or posting messages to the students.

Inquiries can be send to Kenneth Oksanen, when he or some of the assistants will reply in few days notice. Consider asking only course organisational matters by e-mail. With other problems (how algorithm works, programming problems, experiment plans, writing the report, ...) students should ask for a meeting with their responsible assistant.

The course will not have physical noticeboard.

8. Enrollment

More specific instructions about enrollment will be published on the course homepage.