Soft Cost Consensus Models In GDM And Their Applications In Loan Consensus Reaching

Group decision-making(GDM)aims to solve unstructured decision problems that require the participation of experts or decision makers.The decision group often needs to conduct extensive discussions,and constantly modify their opinions and preferences to form a consensus that is generally accepted by members of the group.The formation of consensus is complex,which consumes a lot of manpower and material resources.How to reach a consensus under limited resources is a hot research problem in GDM.This dissertation mainly studies two important issues to reach consensus.First,the optimization of consensus cost.Reaching a consensus is complicated and resourceconsuming,the issue of consensus cost is one of the important factors that needs to be considered.In actual decision process,to reach a consensus requires a lot of resources to persuade individuals to change their opinions.Therefore,the consensus cost needs to be optimized under limited resource to effectively reach a consensus.Second,a moderator's preference on consensus level.A consensus process often requires an important individual or organization(moderator)with a leadership and coordinating role to guide the consensus process.The participation of a moderator can reduce uncertainty and improve the efficiency of consensus reaching.The moderator plays an important role in GDM process,and his/her preference on consensus level affects the consensus cost and the changes of experts' opinions.However,the existing cost consensus methods do not consider this important factor.Based on the group decision-making theory and the multiobjective optimization theory,this dissertation conducts research on the above problems.The main research content and results of this dissertation are as follows:(1)The concept of soft cost consensus is proposed,and a generalized cost consensus model is constructed by defining the consensus degree function and the generalized aggregation operator.Based on the generalized model,three types of soft minimum cost consensus models with different weighted average operators are constructed,and the economic significances of soft cost consensus models are discussed.(2)Based on the Dual Theory of Multi-objective Programming,a maximum compensation consensus model is constructed from the perspective of experts.The economic significance of the model is analyzed from the variables,constraints and objective function of the model.Based on this,the relationships between the soft minimum cost model and the maximum compensation model are discussed.The Relationships among the expected unit compensation of an expert,the compositions of an expert's expected unit compensation and his/her adjusted opinion are studied from the theoretical perspective.(3)This dissertation introduces the concept of utility to model moderator' attitude towards the consensus level,and constructs three types of consensus optimization models based on moderator's neutral,radical and mixed preferences.The properties and practical significances of the models are studied,and the influences of these three preferences of a moderator on the consensus cost and the opinions of experts are analyzed.(4)This dissertation proposes a soft consensus-reaching framework based on bilateral communication mechanism.The proposed framework considers both sides in consensus process,including the opinions of experts and the preference of a moderator.Therefore,it reflects the bilateral communication and compromise mechanism between the moderator and the experts in consensus process,and provides a more flexible and efficient consensus process for GDM.(5)This dissertation applies the soft cost consensus method to the funds matching problem of online lending,and constructs soft loan consensus models under a certain consensus level.The effectiveness of the proposed models are verified by analyzing the influence of moderator's(the online lending platform)preference on lender's rate of return and total consensus cost under different consensus levels.Meanwhile,it provides a new way to coordinate the funding supply and demand for large-scale peer-to-peer lending,and further expands the application of soft cost consensus method.The contributions of this research are two-folds: Theoretically,this dissertation extends the minimum cost consensus to the level of soft consensus,which makes up for the deficiencies of soft consensus cost research.It reveals the compensation mechanism for moderator and experts to reach soft consensus from the theoretical level,which helps to promote the consensus of GDM.The study of moderator's preference on consensus level is important for exploring the influence of a moderator on the consensus result,and revealing the role of a moderator in the consensus-reaching mechanism.It makes up for the lack of research on the moderator's preference on consensus level in GDM.In terms of application,the proposed models can be applied to the loan consensus of online lending.It benefits not only for the lending platform to reduce the probability of losing bids,but also helps to increase the investor's rate of return.It can help platforms to increase the efficiency of funds matching,which has certain practical significance.