考虑情境依赖决策行为的多准则决策偏好学习方法

doi:10.16381/j.cnki.issn1003-207x.2024.1896

Abstract

Abstract:

Multi-criteria decision analysis （MCDA） often requires decision makers （DMs） to provide preference information for a subset of reference alternatives， thereby reducing their cognitive burden while capturing their underlying value functions. When a DM’s holistic preference information is inconsistent with an assumed model， either some assignments cannot be reproduced， or the DM must revise their preferences. In this study， a specific challenge is addressed： the inconsistency arising from criterion context-dependent decision behavior. As DMs gather more information about alternatives， they may adopt different evaluation strategies across various criteria contexts， leading to inconsistencies in their preference information. To tackle this issue， the DM’s preferences are modeled using a piecewise linear value function and develop a two-stage methodological framework. In the first stage， an optimization model is constructed to assess the consistency of the provided preference information and to determine the number of value functions required for an accurate reconstruction. In the second stage， leveraging the principle of parsimony， an optimization model is introduced to select a complementary set of value functions that best reconstruct the DM’s preference information. Furthermore， classification and regression tree （CART） algorithms are integrated to extract rules that correlate different criteria contexts with the corresponding value functions， thereby enhancing the interpretability and decision support of the model. The proposed approach is validated through a comprehensive case study on the evaluation of research units. Detailed numerical experiments， comparative assessments， and simulation analyses demonstrate that our method effectively captures the context-dependent decision behavior of DMs， while also improving the robustness and explanatory power of the preference model.

Key words: multiple-criteria decision-making, context-dependent decision behavior, value function, preference learning

CLC Number:

C934

Qian Liang,Zhongquan Hu,Zhen Zhang, et al. Multi-Criteria Preference Learning Method Considering Context-Dependent Decision Behavior[J]. Chinese Journal of Management Science, 2026, 34(9): 79-89.

Figures/Tables 12

准则	含义
科研活动 $(g 1)$	包括期刊论文、专著、专利和设计等，该项准则值反映了该领域内的研究人员发表文章或者发明所对应的平均得分。
科研潜力 $(g 2)$	包括授予该领域研究人员的学位数量，研究成员的流动性，实验室的认证，对学科项目发展的贡献，国际科学组织或期刊编辑委员会的成员资格。通过对上述的指标进行综合评价后得到该项准则值。
物质影响 $(g 3)$	指正在进行的项目和业界合作的财务方面。例如，该领域通过技术发明、软件实施、算法设计或研究成果商业化所获得的收入。
非物质影响 $(g 4)$	指专家对该领域内的十大最重要的成就评估。例如，对社会有重要影响的研究成果、科学成果奖、组织国际会议等。

类别	方案
参考方案集	$a 35, a 28, a 47, a 42, a 43, a 21, a 24, a 53, a 6, a 5, a 48, a 45, a 10, a 44, a 34, a 1, a 22, a 37, a 56, a 30, a 9, a 26, a 20, a 49, a 36, a 17, a 2, a 8, a 57, a 18, a 29, a 27, a 15, a 3, a 46, a 38, a 39, a 31, a 12, a 50 .$
非参考方案集	$a 41, a 14, a 16, a 33, a 52, a 40, a 4, a 7, a 13, a 32, a 11, a 25, a 55, a 19, a 23, a 54, a 51 .$

方法	子区间数	$g 1$	$g 2$	$g 3$	$g 4$
UTA^[37]	$γ j = 2$	0.65	0.11	0.12	0.12
	$γ j = 3$	0.98	0.01	0.01	0.00
	$γ j = 4$	0.99	0.00	0.01	0.00
	$γ j = 5$	0.96	0.01	0.03	0.00
模型 (9)	$γ j = 2$	[0.60; 0.00; 0.54]	[0.05; 0.01; 0.00]	[0.21; 0.97; 0.46]	[0.14; 0.02; 0.00]
	$γ j = 3$	[0.48; 0.87]	[0.32; 0.00]	[0.18; 0.11]	[0.02; 0.02]
	$γ j = 4$	[0.47; 0.21]	[0.27; 0.00]	[0.19; 0.61]	[0.07; 0.18]
	$γ j = 5$	[0.50; 0.93]	[0.12; 0.00]	[0.31; 0.07]	[0.07; 0.00]
模型 (10)	$γ j = 2$	[0.51; 0.14; 0.19]	[0.07; 0.54; 0.00]	[0.29; 0.20; 0.75]	[0.13; 0.12; 0.06]
	$γ j = 3$	[0.41; 0.15]	[0.30; 0.06]	[0.27; 0.70]	[0.02; 0.09]
	$γ j = 4$	[0.50; 0.10]	[0.25; 0.04]	[0.18; 0.76]	[0.07; 0.10]
	$γ j = 5$	[0.45; 0.13]	[0.22; 0.02]	[0.29; 0.79]	[0.04; 0.06]

方法	子区间数
方法	$γ j = 2$	$γ j = 3$	$γ j = 4$	$γ j = 5$
UTA	$0.0559 ± 0.0462$	$0.0520 ± 0.0409$	$0.0574 ± 0.0458$	$0.0571 ± 0.0452$
模型(9)	$0.0595 ± 0.0459$	$0.0458 ± 0.0401$	$0.0495 ± 0.0400$	$0.0470 ± 0.0423$
模型(10)	$0.0513 ± 0.0385$	$0.0427 ± 0.0332$	$0.0447 ± 0.0330$	$0.0446 ± 0.0321$

对比方法	$γ j = 2$	$γ j = 3$	$γ j = 4$	$γ j = 5$
模型(9)-UTA	$0.995$	$0 . 000 *$	$0 . 000 *$	$0 . 000 *$
模型(10)-UTA	$0 . 000 *$	$0 . 000 *$	$0 . 000 *$	$0 . 000 *$
模型(10)-模型(9)	$0 . 000 *$	$0 . 000 *$	$0 . 000 *$	$0 . 008 *$

References 40

[1]	Dyer J S， Fishburn P C， Steuer R E， et al. Multiple criteria decision making， multiattribute utility theory： The next ten years［J］. Management Science， 1992， 38（5）： 645-654.
[2]	王浩伦，张发明，完颜晓盼，等. 基于q阶orthopair模糊集和参考理想法的多准则决策方法［J］. 中国管理科学， 2023， 31（10）： 245-253.
	Wang H L， Zhang F M， Wanyan X P， et al. A multi-criteria decision-making method based on q-rung orthopair fuzzy set and reference ideal method［J］. Chinese Journal of Management Science，2023，31（10）： 245-253.
[3]	杨荣庆，唐孝安，张强，等. 分布式乘性偏好环境下考虑决策者偏好调整意愿的最优-最劣多准则决策方法［J］. 中国管理科学， 2024， 32（7）： 65-75.
	Yang R Q， Tang X A， Zhang Q， et al. Best-worst multi-criteria decision-making method with decision maker’s preference adjustment willingness under distributed multiplicative preference environment［J］. Chinese Journal of Management Science， 2024， 32（7）： 65-75.
[4]	Corrente S， Greco S， Kadziński M， et al. Robust ordinal regression in preference learning and ranking［J］. Machine Learning， 2013， 93（2-3）： 381-422.
[5]	Fallah Tehrani A， Cheng W， Dembczyński K， et al. Learning monotone nonlinear models using the Choquet integral［J］. Machine Learning， 2012， 89（1-2）： 183-211.
[6]	杨娜，廖貅武，雷宏振. 考虑属性关联的两阶段获胜者确定模型［J］.中国管理科学， 2019， 27（10）： 148-158.
	Yang N， Liao X W， Lei H Z. Two stage winner determination approach with interacting attributes［J］. Chinese Journal of Management Science， 2019， 27（10）： 148-158.
[7]	Liu J， Kadziński M， Liao X， et al. Data-driven preference learning methods for value-driven multiple criteria sorting with interacting criteria［J］. INFORMS Journal on Computing， 2020，33（2）： 419-835.
[8]	Ghaderi M， Ruiz F， Agell N. A linear programming approach for learning non-monotonic additive value functions in multiple criteria decision aiding［J］. European Journal of Operational Research， 2017， 259（3）： 1073-1084.
[9]	Liu J， Liao X， Kadziński M， et al. Preference disaggregation within the regularization framework for sorting problems with multiple potentially non-monotonic criteria［J］. European Journal of Operational Research， 2019， 276（3）： 1071-1089.
[10]	Kadziński M， Martyn K， Cinelli M， et al. Preference disaggregation for multiple criteria sorting with partial monotonicity constraints： Application to exposure management of nanomaterials［J］. International Journal of Approximate Reasoning， 2020， 117： 60-80.
[11]	Li Z， Zhang Z， Pedrycz W. Integrating machine learning models to learn potentially non-monotonic preferences for multi-criteria sorting from large-scale assignment examples［J］. Omega， 2025， 131： 103219.
[12]	Zhang Z， Li Z， Yu W. Lexicographic optimization-based approaches to learning a representative model for multi-criteria sorting with non-monotonic criteria［J］. Computers & Operations Research， 2025， 175： 106917.
[13]	Aggarwal M， Fallah Tehrani A. Modelling human decision behaviour with preference learning［J］. INFORMS Journal on Computing， 2019， 31（2）： 318-334.
[14]	Kadziński M， Ghaderi M， Dąbrowski M. Contingent preference disaggregation model for multiple criteria sorting problem［J］. European Journal of Operational Research， 2020， 281（2）： 369-387.
[15]	Wen Z， Liao H， Figueira J R. A preference disaggregation-driven multiple criteria sorting model based on regret theory［J］.Omega， 2024， 129： 103150.
[16]	Gehrlein J， Miebs G， Brunelli M， et al. An active preference learning approach to aid the selection of validators in blockchain environments［J］. Omega， 2023， 118： 102869.
[17]	Hüllermeier E， Słowiński R. Preference learning and multiple criteria decision aiding： Differences， commonalities， and synergies： Part II［J］. 4OR， 2024， 22（3）： 313-349.
[18]	Mousseau V， Figueira J， Dias L， et al. Resolving inconsistencies among constraints on the parameters of an MCDA model［J］. European Journal of Operational Research， 2003， 147（1）： 72-93.
[19]	Liu J， Liao X， Huang W， et al. Market segmentation： A multiple criteria approach combining preference analysis and segmentation decision［J］. Omega， 2019， 83： 1-13.
[20]	Guo M， Liao X， Liu J. A progressive sorting approach for multiple criteria decision aiding in the presence of non-monotonic preferences［J］. Expert Systems with Applications， 2019， 123： 1-17.
[21]	Angilella S， Greco S， Matarazzo B. Non-additive robust ordinal regression： A multiple criteria decision model based on the Choquet integral［J］. European Journal of Operational Research， 2010， 201（1）： 277-288.
[22]	Tversky A， Simonson I. Context-dependent preferences［J］. Management Science， 1993， 39（10）： 1179-1189.
[23]	Kivetz R， Netzer O， Srinivasan V. Extending compromise effect models to complex buying situations and other context effects［J］.Journal of Marketing Research， 2004， 41（3）： 262-268.
[24]	Zheng Q， Pan X A， Carrillo J E. Probabilistic selling for vertically differentiated products with salient thinkers［J］. Marketing Science， 2019， 38（3）： 442-460.
[25]	Dzyabura D， Hauser J R. Recommending products when consumers learn their preference weights［J］. Marketing Science， 2019， 38（3）： 417-441.
[26]	Liu J， Kadziński M， Liao X. Modeling contingent decision behavior： A Bayesian nonparametric preference-learning approach［J］. INFORMS Journal on Computing， 2023， 35（4）： 764-785.
[27]	Gao P， Ma Y， Chen N， et al. Assortment optimization and pricing under the multinomial logit model with impatient customers： Sequential recommendation and selection［J］. Operations Research， 2021， 69（5）： 1509-1532.
[28]	Montibeller G， von Winterfeldt D. Cognitive and motivational biases in decision and risk analysis［J］. Risk Analysis， 2015， 35（7）： 1230-1251.
[29]	Martyn K， Kadziński M. Deep preference learning for multiple criteria decision analysis［J］. European Journal of Operational Research， 2023， 305（2）： 781-805.
[30]	Greco S， Słowiński R， Wallenius J. Fifty years of multiple criteria decision analysis： From classical methods to robust ordinal regression［J］. European Journal of Operational Research， 2025， 323（2）： 351-377.
[31]	Zhou K， Gong Z， Wei G， et al. Preference disaggregation analysis with criteria selection in a regularization framework［J］. Omega， 2025， 133： 103252.
[32]	Li Z， Zhang Z， Pedrycz W. An incremental preference elicitation-based approach to learning potentially non-monotonic preferences in multi-criteria sorting［J］. European Journal of Operational Research， 2025， 323（2）： 553-570.
[33]	Keeney R， Raiffa H. Decisions with multiple objectives： Preferences and value tradeoffs［M］. New York： John Wiley & Sons， 1976.
[34]	Kadziński M， Wójcik M， Ciomek K. Review and experimental comparison of ranking and choice procedures for constructing a univocal recommendation in a preference disaggregation setting［J］. Omega， 2022， 113： 102715.
[35]	Lazar N. Ockham’s razor［J］. WIREs Computational Statistics， 2010， 2（2）： 243-246.
[36]	Gordon A D， Breiman L， Friedman J H， et al. Classification and regression trees［J］. Biometrics， 1984， 40（3）： 874.
[37]	Greco S， Mousseau V， Słowiński R. Ordinal regression revisited： Multiple criteria ranking using a set of additive value functions［J］. European Journal of Operational Research， 2008， 191（2）： 416-436.
[38]	Kadziński M， Ghaderi M， Wąsikowski J， et al. Expressiveness and robustness measures for the evaluation of an additive value function in multiple criteria preference disaggregation methods： An experimental analysis［J］. Computers & Operations Research， 2017， 87： 146-164.
[39]	徐选华，马志鹏，陈晓红. 基于公众偏好大数据分析的大群体应急决策质量动态演化研究［J］. 中国管理科学， 2022， 30（7）： 140-149.
	Xu X H， Ma Z P， Chen X H. Dynamic evolution research on emergency decision quality of large group based on the public preferences big data［J］. Chinese Journal of Management Science， 2022， 30（7）： 140-149.
[40]	缑迅杰，邓富民，徐泽水. 基于自信双层语言偏好关系的大规模群体共识决策方法及其应用研究［J］. 中国管理科学， 2023， 31（9）： 222-232.
	Gou X J， Deng F M， Xu Z S. The application of the large-scale group consensus decision-making method based on self-confident double hierarchy linguistic preference relations［J］. Chinese Journal of Management Science， 2023， 31（9）： 222-232.

文献	决策问题	是否考虑效用函数的复杂度	是否考虑情境依赖的行为因素
[13]	多准则排序	否	否
[14]	多准则分类	否	是
[15]	多准则分类	否	否
本文	多准则排序	是	是

Multi-Criteria Preference Learning Method Considering Context-Dependent Decision Behavior

RichHTML

PDF (PC)

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 12

References 40

Related Articles 4

Metrics

Comments

Recommended 0

[1]	Xi Chen, Youqi Dou, Wenbo Zhang. An Intelligent Auxiliary Diagnosis Method for Diseases Integrating Preference Learning and Evidential Reasoning [J]. Chinese Journal of Management Science, 2026, 34(7): 177-188.
[2]	LUO Biao, WANG Cheng-yuan. Design and Analysis of Compensation Contract Model：Based on the Cumulative Prospect Theory [J]. Chinese Journal of Management Science, 2013, 21(5): 121-128.
[3]	WEN Feng-hua, RAO Gui-tian, YANG Xiao-guang. Empirical Study on Value Functions of Stock Markets [J]. Chinese Journal of Management Science, 2010, 18(5): 7-13.
[4]	WU Jian. A New Approach for Priorities Trapezoidal Fuzzy Number Reciprocal Judgement Matrix [J]. Chinese Journal of Management Science, 2010, 18(3): 95-100.