TY - JOUR
T1 - An evidential dynamical model to predict the interference effect of categorization on decision making results
AU - He, Zichang
AU - Jiang, Wen
N1 - Publisher Copyright:
© 2018 Elsevier B.V.
PY - 2018/6/15
Y1 - 2018/6/15
N2 - Categorization is necessary for many decision making tasks. However, the categorization process may interfere the decision making result and bring about the disjunction fallacy. To predict the interference effect of categorization, some models based on quantum cognition theory have been proposed. In quantum dynamical models, like the quantum belief-action entanglement (BAE) model, actions and beliefs are deemed to be entangled. However, the entanglement degree is an artificially defined parameter. In this paper, a new evidential dynamical (ED) model based on Dempster–Shafer (D-S) evidence theory and quantum dynamical modelling is proposed. Considering that sometimes people hesitate to make a decision, it is reasonable to extend the action states by introducing an uncertain state. In an evidential framework, categorization can influence the uncertain state in actions. The interference effect is measured by handling the uncertain state while no extra parameter is defined artificially. The proposed model is applied to the classical categorization decision-making experiments. Compared with the existing models, the number of free parameters in the ED model is less than the classical quantum models, and the ED model is more rational and simpler than an evidential Markov model. The model application results and discussions show the correctness and effectiveness of the ED model. Not only the interference effect of categorization on decision making results is explained and predicted, but also an inspiring dynamical decision making framework is proposed in this paper. We believe that the proposed ED model will bring more opportunities and will result in more applications in the future.
AB - Categorization is necessary for many decision making tasks. However, the categorization process may interfere the decision making result and bring about the disjunction fallacy. To predict the interference effect of categorization, some models based on quantum cognition theory have been proposed. In quantum dynamical models, like the quantum belief-action entanglement (BAE) model, actions and beliefs are deemed to be entangled. However, the entanglement degree is an artificially defined parameter. In this paper, a new evidential dynamical (ED) model based on Dempster–Shafer (D-S) evidence theory and quantum dynamical modelling is proposed. Considering that sometimes people hesitate to make a decision, it is reasonable to extend the action states by introducing an uncertain state. In an evidential framework, categorization can influence the uncertain state in actions. The interference effect is measured by handling the uncertain state while no extra parameter is defined artificially. The proposed model is applied to the classical categorization decision-making experiments. Compared with the existing models, the number of free parameters in the ED model is less than the classical quantum models, and the ED model is more rational and simpler than an evidential Markov model. The model application results and discussions show the correctness and effectiveness of the ED model. Not only the interference effect of categorization on decision making results is explained and predicted, but also an inspiring dynamical decision making framework is proposed in this paper. We believe that the proposed ED model will bring more opportunities and will result in more applications in the future.
KW - Categorization decision-making experiment
KW - Dempster–Shafer evidence theory
KW - Disjunction fallacy
KW - Evidential dynamical model
KW - Interference effect of categorization
KW - Quantum dynamical model
UR - http://www.scopus.com/inward/record.url?scp=85044003932&partnerID=8YFLogxK
U2 - 10.1016/j.knosys.2018.03.014
DO - 10.1016/j.knosys.2018.03.014
M3 - 文章
AN - SCOPUS:85044003932
SN - 0950-7051
VL - 150
SP - 139
EP - 149
JO - Knowledge-Based Systems
JF - Knowledge-Based Systems
ER -