Considering undesirable variables in PCA-DEA method: a case of road safety evaluation in Iran

Subject Areas : Mathematical Optimization

Alimohammad Ahmadvand ¹ , Zeinab Abtahy ² , Mahdi Bashiri ³

1 - Professor, Dept. of Industrial Engineering, Emam Hossein University, Tehran, Iran
2 - M.Sc., Dept. of Industrial Engineering, Khoramabad Branch, Islamic Azad University, Khoramabad, Iran
3 - Associate Professor, Dept. of Industrial Engineering, Shahed University, Tehran, Iran

Received: 2015-05-17 Accepted : 2015-05-17 Published : 2011-11-01

Keywords: DEA, PCA, PCA-DEA, Performance Evaluation, road safety, undesirable variables,

Abstract :

This paper presents a deterministic approach for performance assessment of different province’s road safety level at Iran. A data envelopment analysis (DEA) model considering undesirable input and output indices and a multivariate statistical method, principle component analysis (PCA) are used in this paper, while previous studies do not use composite PCA-DEA method and undesirable input and output indices simultaneously. PCA is used to improve discrimination power of DEA and making variables as independent as possible to avoid overlapping of decision making units (DMU’s) information. The advantage of the proposed approach is benchmarking the most unsafe roads to allocate bounded road safety public budget in the best manner while considering undesirable output and input variables simultaneously.