Incorporating value judgment in Technology Forecasting using Data Envelopment Analysis

Abstract

Technology Forecasting using Data Envelopment Analysis (TFDEA) has been employed to a wide range of applications because of its ability to model complex tradeoffs. The very name indicates that it is based on Data Envelopment Analysis (DEA) which has strength that it doesn't require fixed a priori weighting scheme. Instead, it adopts dynamic weighting scheme that each data point can choose their best possible weights. However, it is well known that this flexibility may result in extreme weights that may be considered unrealistic in certain applications and has been one subject of DEA researches. This paper extends the standard TFDEA model to incorporate value judgment in assessment to refine the analysis framework. The proposed model is applied to the Liquid Crystal Display (LCD) industry to address impact of various weight restrictions on the technology forecast results