Applying ant systems to two real-life assignment problems

Cancela, Héctor

Resumen:

Ant Systems (AS) is a recently proposed meta-heuristic inspired on biological behaviors, which has been applied to a variety of combinatorial optimization problems, including the QAP (Quadratic Assignment Problems). In this work, we have studied the adaptation of the Ant Systems meta-heuristic to two different real-life assignment problems, which appear in educational institutions: the timetabling problem (assigning courses to classrooms and times), and the assignment of final proyects to students. There is no standard definition for these problems, as in each institution the rules and objectives are different. We have been successful in adapting. AS to tackle these problems as defined by our institution rules, showing the adaptability of this meta-heuristic to complex, real-life problems. In both cases the AS meta-heuristic obtained good quality solutions (in the timetabling case, at the cost of longer running times).


Detalles Bibliográficos
2000
COMBINATORIAL OPTIMIZATION
METAHEURISTICS
ASSIGNMENT
OPTIMIZACION COMBINATORIA
Universidad de la República
COLIBRI
http://hdl.handle.net/20.500.12008/3513
Acceso abierto
Licencia Creative Commons Atribución – No Comercial – Sin Derivadas (CC BY-NC-ND 4.0)
Resumen:
Sumario:Ant Systems (AS) is a recently proposed meta-heuristic inspired on biological behaviors, which has been applied to a variety of combinatorial optimization problems, including the QAP (Quadratic Assignment Problems). In this work, we have studied the adaptation of the Ant Systems meta-heuristic to two different real-life assignment problems, which appear in educational institutions: the timetabling problem (assigning courses to classrooms and times), and the assignment of final proyects to students. There is no standard definition for these problems, as in each institution the rules and objectives are different. We have been successful in adapting. AS to tackle these problems as defined by our institution rules, showing the adaptability of this meta-heuristic to complex, real-life problems. In both cases the AS meta-heuristic obtained good quality solutions (in the timetabling case, at the cost of longer running times).