The value efficiency approach is one possible way of incorporating information preferences into performance analysis of Decision-Making Units (DMUs). In this paper, we propose a novel geometric interpretation of value efficiency while plugging it into radial and non-radial DEA (Data Envelopment Analysis) models under the assumption of variable returns to scale. In addition, this novel geometric interpretation of value efficiency is extended to Additive Slacks-Based Measure (ASBM) modeling. This is achieved by linearization of non-radial DEA models using multi-objective programming. Performance of such proposed approaches in terms of reliability and discriminatory power are compared through a case study involving Finish bank branches. Research implications are then derived and conclusions are drawn.

价值效率方法是将信息偏好纳入决策单元 (DMU) 性能分析的一种可能方法。在本文中，我们提出了一种新的价值效率几何解释，同时在规模收益可变的假设下将其插入径向和非径向 DEA（数据包络分析）模型。此外，这种对价值效率的新颖几何解释被扩展到基于附加松弛的度量 (ASBM) 建模。这是通过使用多目标规划对非径向 DEA 模型进行线性化来实现的。通过涉及芬兰银行分行的案例研究，比较了这些建议方法在可靠性和辨别力方面的性能。然后得出研究意义并得出结论。