Requirements and Evaluation

The project will be presented during the exams.  Students will be graded on the quality of their project, their article, the presentation, their reviews and their participation during the exam seminar.

Students will write down their work in an article format (, for which the LaTeX style is provided on the website. (see formatting)

An essential constraint is that the article should contain sufficient information to reproduce the results. So be as clear and specific as possible when you describe the model and its parameters.   Also provide a clear and logical interpretation of your results.

There are additional points to be gained for originality, creativity and scientific relevance.   We hence encourage each group to go beyond the specific information provided for each topic and demonstrate their interest in this course.

The article should consist of at least 5 pages (see formatting) and will be handed in at least two weeks before the exam (see dates below). 

Upon receiving the article it will be distributed to another group who will have to review the article.  Reviewing entails that one writes a document discussing the following issues (2-3 pages):

1) Does the introduction explain clearly the content of the paper

2) whether there is sufficient background information to understand the relevance of the work

3) whether the methods are clearly explained (can the results be reproduced?)

4) whether the results answer the questions asked in the paper.

5) whether all questions are answered

6) whether the conclusion is sufficient

7) and whether the overall style is ok and

8) whether you believe things are missing in the discussion.

9) etc.

We also expect that you add to this review:

1) 3 positive points concerning the work, clearly specifying why you think they are well-done or interesting

2) 3 negative points, which may include missing/unclear explanations or suggestions for improvement

3. 3)at least 3 clear and relevant questions on the content or the methods used which can be asked (next to other questions).
   

The review has to be mailed no later than the 13 of January 2020 to tlenaert@ulb.ac.be

The working language is English.   Make sure that spelling and grammar mistakes are minimal.         

Note that plagiarism is not tolerated.  Each paper will be analyzed with Compilatio.net.  When the analysis is green no actions will be taken, but in the orange or red case, we are obliged to take  action and exclude you from the exam session.

   

Formatting

Structure of the article: (Introduction, Methods, results, discussion, references)

1. 1.In the introduction a clear specification of the problem is provided, including related work. Look for additional but relevant references which you can cite in this part.
   

2. 2.In the Methods section you specify how the model is constructed
   

3. 3.In the Results section you provide a descriptions of the simulations you have done and their results.  Also include the information concerning the parameter settings.
   

4. 4.In the Discussion section you provide a summary and an explanation of your work.
   

5. 5.A good scientific article has a bibliography
   

6. 6.
   

The following article provides an example of how an article should be structured : example.pdf

We will use the Alife XII formatting template for each article.  Below you can find a pdf explaining the formatting instructions.  Next you find a zip-file containing an example (and style files) to write your article in LaTeX.  For those who prefer we also added the style file for word.  Please make use of the bibliographic tools of Word when you use this document

The final manuscript should be provided in PDF and send to tlenaert@ulb.ac.be.  Ignore the submission explanation in the formatting instructions.

formatting_instructions.pdf

alife-example.tar.gz

ALifeXII_Word_template.doc

Project proposals

Organisation

The project will allow students to experiment with the principles of learning discussed during the course.                                

The project will be performed in groups of 4 students.

We provide here a number of proposals but students are also allowed to suggest their own proposal, which will be evaluated by the academic staff.

Each group will send an email to Prof. Tom Lenaerts containing a ranking of 3 projects according to their preferences (by December 6, 2019 no later than 1PM).  A project will be assigned which will take into account these preferences. 

The first set of proposals asks you to read an article, write a software that reproduces the results discussed in the paper.  Additional credits can be gained for extending the work discussed in those papers in a new and innovative manner. Describe in the manuscript  the context (search for additional articles), the methodology YOU used to arrive at the results and extensions and YOUR OWN explanation of the results you observe and how they relate to the results in those articles.  Propose also possible extensions to the work you produced in the Discussion section.

These articles can be found through google scholar, if not contact either Prof. T. Lenaerts or Prof. A. Nowé

dummy article : Littman, M. L. (1994). Markov games as a framework for multi-agent reinforcement learning. In Machine learning proceedings 1994 (pp. 157-163). Morgan Kaufmann. (https://www2.cs.duke.edu/courses/spring07/cps296.3/littman94markov.pdf)

Important dates

Success!

For this project, there are two steps: a preparation and the actual exam. During the exam, the work is presented and evaluated by your fellow students and the academic staff.

Choice : December 6, 2019 1PM (email to tlenaert@ulb.ac.be)

Article submission : January 7, 2020, noon (email to tlenaert@ulb.ac.be)

Review submission: January 13, 2020, noon (email to tlenaert@ulb.ac.be)

Date exam : see schedule.