This is an old webpage! For the spring 2014 course, visit: http://www.cse.chalmers.se/edu/course/TIN172/

Artificial Intelligence, LP4, VT2013

The Course Projects

Your mission, should you decide to accept it, is to explore strange new worlds, to seek out new life and new civilizations, to boldly go where no one has gone before.

The projects are the heart of the course. Typically, you set out to solve some problems or to explore an area of AI, and build a system or a suite of programs to do this. The problems or areas will probably not be new, but they are neither completely researched.

In each group you will implement two projects:

  • One project where you will use data-driven methods such as machine learning and/or statistics.
  • A second project where you will implement a system that uses symbolic methods such as planning and/or logical inference.

Your programs may start from scratch, or build on existing available code and tools. In either case, your report must compare your programs with other possible approaches. It must explain the key parts of your programs using pseudo-code, and give your own understanding of the area with examples (at least some of which should be your own). It should also show an understanding of the relevant literature, referenced, and show how your work fits into the generally accepted picture.

You will work in teams of four students each, and submit all your work as a team. You will also maintain a diary, updated weekly, which records briefly who did what that week. Show your diary to your supervisor during each supervision session and include it in your final report.

Fire

All submissions should be done via the Fire system. The first thing you have to do is to register your group. Use the link in the menu to the left.

Project report

The report should not be too long, and not too short… I guess around 10 pages (excluding appendix) is reasonable, but that depends on a lot of things, both external (font family, font size, columns, row spacing, figures, …) and internal (writing style, explanation detail, content, …).

There is a very simple template for the project report, which you can use as a first baseline.

Supervision

Note: When mailing your supervisor, please use a subject that starts with [AI] — this makes things much easier for us!

You have supervision sessions all course weeks, except the first.

These are the main teaching occasions, admittedly only once a week for 40 minutes, but this is quality time, as there are only 4 students in the room. These sessions are supplemented by feedback on your submissions.

We suggest you bring a laptop to every session. Show the supervisor the status of your project: the diary, reports, code documentation, and demonstrate your running code. Then you can discuss your achievements and your problems, and make a plan for the forthcoming week. One of the students should take minutes, both to ensure you don't forget what was said, and to give you a record against which you can compare your progress.

The supervisor's main role is that of research coach. They cannot be expected to answer all your questions, since they only spend about two hours a week on your project, whereas you as a team spend 80. You are expected to be the experts on your topic, while the supervisor's expertise is in the business of exploring the scientific world.