Abstract
This paper deals with a special case of multicriteria optimization problems. The problems studied come from the medical domain and are of a very important practical relevance. One of the problems refers to the ranking of treatments for the Trigeminal Neuralgia. The second problem refers to a hierarchy of risk factors for Bronchial Asthma. The most common way to deal with a multiobjective optimization problem is to apply Pareto dominance relationship between solutions. But in the cases studied here, a decision cannot be made just by using Pareto dominance. In one of the experiments, all the potential solutions are nondominated (and we need to clearly find a hierarchy of these solutions) and in the second experiment most of the solutions are nondominated between them. We propose a novel multiple criteria procedure and then an evolutionary scheme is applied for solving the problems. Results obtained by the proposed approach in a very simple way are same as the results (or even better) obtained by applying weighted-sum method. The advantage of the proposed technique is that it does not require any additional information about the problem (like weights for each criteria in the case of weighted-sumapproach).
| Original language | English |
|---|---|
| Pages (from-to) | 1407-1417 |
| Number of pages | 11 |
| Journal | Applied Soft Computing Journal |
| Volume | 8 |
| Issue number | 4 |
| DOIs | |
| Publication status | Published - Sept 2008 |
Keywords
- Bronchial asthma
- Evolutionary algorithm
- Medical diagnosis
- Multicriteria analysis
- Pareto dominance
- Trigeminal neuralgia
- Weighted sum method