Reaching heterogeneous group consensus with classification error for sorting medical emerging technology service providers

Yuanyuan Fu, Decui Liang, Alessio Ishizaka, Dengfeng Li

Abstract: In order to improve diversity services and boost diagnostic accuracy, more and more emerging technologies are applied to assist medical treatment. The medical emerging technology service provider selection is a noteworthy problem for hospital risk management. Thus, we develop a heterogeneous group AHPSort (GAHPsort) to provide the recommendation for hospitals and explore the consensus of GAHPSort to avoid the conflict. Specifically, by considering the difference of decision makers' evaluation standard, we design a group sorting consensus of GAHPSort from the angle of sorting result via pairwise comparison of decision makers. Meanwhile, we construct asymmetrical classification error cost function to describe classification error costs of different decision makers' preferences. Further, a minimum adjustment model based on sorting consensus index with classification error is established. Considering that the results of technology providers in the same class are close, this paper develops the distinction between the inter-class and the between-class distances to depict the weak opinion consensus and extend the minimum adjustment model. Besides, in order to avoid the unexplainable evaluation, we design the semantic tolerance and further improve the minimum adjustment model. Finally, we utilize two cases to validate our proposed method via comparison analysis.