题目： Data Envelopment Analysis for Effectiveness Measure and Marginal Abatement Cost Estimation in the Power System Industry
主讲人：Dr. Chia-Yen Lee （National Cheng Kung University）
时间：2016年7月12日14:00
地点：主楼六层会议室
主讲人介绍：

    Dr. Chia-Yen Lee is an associate professor in the Institute of Manufacturing Information and Systems, National Cheng Kung University, Taiwan. He received the BS in Mathematical Sciences and BBA in Management Information Systems in National Chengchi University, Taiwan; the MS in industrial engineering at National Tsing Hua University, Taiwan; and Ph.D. degree in industrial and systems engineering at Texas A&M University, USA. He was an industrial engineer in taiwan semiconductor manufacturing company (tsmc). He currently serves as Associate Editor for Flexible Services and Manufacturing Journal and Co-Principal Investigator at Semiconductor Technologies Empowerment Partners Consortium for Big Data Analytics and Optimization Technologies. His research interests include productivity analysis, data science, intelligent manufacturing systems, and stochastic optimization with applications to the following fields: Energy and Pollutant、Semiconductor Manufacturing、TFT-LCD Manufacturing。

内容介绍：

    Data envelopment analysis (DEA) is a nonparametric technique to estimate production function and technical efficiency of each firm based on the input and output factors. However, demand fluctuation may affect the output level and bias the efficiency analysis when units sold are used at an output measure. This talk defines the concept of effective production and proposes an effectiveness measure to capture the demand fluctuations and sales effect on performance measurement by demand-truncated production function. Effectiveness complements typical efficiency measure. Efficiency measures the relative return on inputs used while effectiveness indicates the ability to match sales given an existing production technology. In addition, the talk also illustrates the estimation of marginal abatement cost (MAC) of undesirable output based on the directional marginal productivity (DMP). The method can avoid estimating MAC of each by-product separately which may lead to an underestimation of the MAC. An empirical study of coal-fired power system is conducted to validate the effectiveness and MAC. We conclude that the proposed effectiveness measure’s ability to distinguish sales and regulation effects from typical productive efficiency, and the DMP also corrects the underestimation issue of the MAC.

（承办：管理工程系，科研与学术交流中心）