Return To Scale Of Integer-Valued Data In Data Envelopment Analysis

Envelopment Analysis (DEA) for decision-making units (DMUs) with integer-valued data While most DEA models assume real-valued data, this study addresses the practical importance of integer data in real-world applications. The research specifically examines two scenarios: ۱. Single integer-valued input and single real-valued output. ۲. Mixed inputs (integer and non-integer) with real-valued outputs. The study introduces the concept of neighborship to extend the RTS algorithm for integer data. It proposes models to calculate left and right returns to scale for each unit in these scenarios. Using both hypothetical and real-world examples, the paper demonstrates the differences between the RTS results from the proposed method and the classical DEA approach. The findings highlight the importance of considering integer data in DEA models for more accurate performance evaluation.

کلیدواژه ها:

Data Envelopment Analysis ، left returns to scale ، right returns to scale ، integer-valued data