The Communication between Congestion and Accepting or Rejecting Outliers Units, an Application of DEA

دورافتاده ها مجموعه اي از واحدهايي هستند كه از واحدهاي ديگر دور مي باشند. قبول يا رد اين واحدها به عوامل مختلفي بستگي دارد. اين مقاله شرايطي را توضيح مي دهد كه دورافتاده ها براي قبول يا رد بدان نيازدارند. مطالعه روي مفهوم تراكم يكي از روشهايي براي اين منظور است. در اين روش ابتدا دور افتاده هايي كه تراكم دارند را تعيين مي كنيم، سپس تصميم به قبول يا رد آنها در مجموعه اي از داده ها را مي گيريم. براي ياري نمودن به بحث مثالهايي ارايه مي شود. علاوه براين بازده به مقياس واحدها يي كه دورافتاده اند بهمراه مثالهايي تعيين مي شوند.

Outliers are considered as a set of data that distinctly stands out from the rest of that data. Accepting or rejecting the outliers depends on various factors. The objective of this paper is to explain the accepting or rejecting conditions of outliers. Studying the congestion of the outlier units is one of the which through which the acceptance or rejection conditions can be figure out. In this method, it is first needed to identify the outliers that have congestion and then decide about the accepting or rejecting them. Discussions are presented following some examples to obtain higher level of underestimating of the proposed method. In addition, the return to scale of outliers are determined and discussed by using some examples.