An evolutionary squeaky wheel optimization approach to personnel scheduling
IEEE Transactions on Evolutionary Computation
Ant colony system: a cooperative learning approach to the traveling salesman problem
IEEE Transactions on Evolutionary Computation
| Hi-index | 0.00 |
This paper proposes an ant colony optimization (ACO) approach to offer online decision support for making admission plans of inpatients. The approach considers patients' severity degrees and urgency levels, aiming to find an admission plan that offers treatment in time to as many patients as possible. At each decision point, the ACO approach builds a construction graph, with each vertices denoting one possible admission time for a patient in wait. Artificial ants walk on the construction graph to construct feasible admission plans by selecting vertices under guides of pheromones and heuristic information. The resulting plans are evaluated from both terms of the total admission rate and the severity degrees of admitted patients. The weights of the two components can be determined according to the preferences of hospital administrators. When implementing the admission plan, only the admissions that are scheduled before the next decision point are actually executed. The rest of the admission plan is used as guides for optimizing the implemented admissions. Simulations based on actual data show that the ACO approach outperforms two classical admission policies and improve the hospital performance in the long run.