https://github.com/franco-ruggeri/course-scheduling
AI algorithms applied to the course scheduling problem
https://github.com/franco-ruggeri/course-scheduling
artificial-intelligence
Last synced: 3 months ago
JSON representation
AI algorithms applied to the course scheduling problem
- Host: GitHub
- URL: https://github.com/franco-ruggeri/course-scheduling
- Owner: franco-ruggeri
- Created: 2020-08-21T10:16:19.000Z (about 5 years ago)
- Default Branch: master
- Last Pushed: 2020-05-07T16:12:23.000Z (over 5 years ago)
- Last Synced: 2025-02-25T19:45:10.091Z (8 months ago)
- Topics: artificial-intelligence
- Homepage:
- Size: 4.38 MB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# AI.A3.PR
## General info
Project of Artificial Intelligence at KTH.
* General topic: advanced planning
* Specific topic: planning with limited resources
* Case study: course scheduling in university
* Algorithms: simulated annealing, genetic algorithm, linear programming
Authors:
* Dingli Mao
* Arturo Rivas Rojas
* Franco Ruggeri
* Ryan Yared
## Structure of the code
Packages:
* default package: contains the Main class to run the test cases
* generator: contains general classes useful to generate the problem and evaluate the solutions
* solvers: contains the implementation of the algorithms
## How to run the test cases
On Windows:
* Import the project into an IDE (e.g. Eclipse, IntelliJ)
* Add the external jars in *src/solvers/lp/* to the build path
+ In Eclipse: right click on the project -> build path -> configure build path -> tab libraries -> add jars
* Run the Main class in one of the following configurations:
+ Generate problems and run GA + Simulated Annealing: set flag *ILP=false*
+ Run ILP on a generated problem: set flag *ILP=true* and *ILP_PROBLEM=* (e.g. *problem_0_0.txt*)
On Ubuntu, you can follow the same steps, but the ILP libraries may give you some problems. In that case, you can anyway run the other algorithms by setting the flag *ILP=false*.
*Remark*: ILP takes a while to finish and in some test cases does not converge. This is the reason why it is run a part and not together with the other algorithms. You should stop manually after a predefined amount of time.
## Output
Problems, solutions and performance will be saved into a subfolder called *output*.
## Customized run
### Generate a new problem
In order to create a new problem, you need a *Generator*. The Generator accepts a range of parameters and creates a valid problem choosing random numbers within those ranges. This can be done using the method *generate()*.
The suggested way is to modify the code in the *Main* class.
### Run a solver
All the algorithms implement the *Solver* interface, that declares just one method: *solve()*. To run a new Solver, just create a new object among the available algorithms and use that method. The algorithms have a set of parameters that can be passed to the constructor.
* Simulated Annealing:
+ Temperature
+ Cooling rate
* Genetic Algorithm:
+ Population size
+ Mutation probability
+ Fitness value that stops the algorithm
+ Maximum time in ms
* Linear Programming:
+ No parameters, the Problem object is enough.