-
Notifications
You must be signed in to change notification settings - Fork 0
1. Introduction
The objective of this research is to solve the following problem statement:
How can you distribute a set of N students over a set of G project groups in an optimal way?
In order to solve this we first have to define the last part of the research question and explain what optimal exactly means in this perspective. This has actually two different parts: pedagogical and computational optimal.
From a pedagogically perspective an optimal way means maximizing learning for each individual student in the group. This boils down in two different questions. First, which students should be assigned to the same group in order to maximize the group project’s learning effectiveness? Second, once a set of criteria has been decided on, how is an optimal group assignment found?
From a computationally perspective this problem is in general an NP-hard [1] problem, finding on optimal solution would involve evaluating all possible assignments, which makes it unfeasible even for small problem sizes. We take a heuristic approach by trying to optimize an objective function. The resulting solution might not be globally optimal, but it should be statistically better than a random assignment.
Our objective is to design and implement a matching algorithm for assigning students to project groups based on various constraints and optimization criteria, e.g. even distribution of skills, group diversity, student preferences, friends and foes, etc. In this wiki you can find more information about the internals of the algorithm, the theory behind it and how specific constraints are modeled in the solver.
The design of our algorithm follows closely the solution proposed by Hübscher [2] with some modifications and extensions. The following sections represent the gist of it.
- E. Ronn, NP-Complete Stable Matching Problems, Computer Science Department, Technion-Israel Institute of Technology, Haifa 3X00, Israel
- R. Hübscher, Assigning Students to Groups Using General and Context-Specific Criteria, IEEE Transactions on Learning Technologies, vol. 3, no. 3, July-September 2010