The TAO Model And Acting Measures Of Three-way Decision Under Granular Computing Perspective

Decision-making is one common behavior in human production and life,including the formulation of policy at the national level to the operation of a machine,i.e.,operational optimization and resource allocation of the industrial field,personalized service and supply chain management of the commercial field,traffic flow control and intelligent navigation of transportation field,and disease diagnosis and treatment strategies of medical field.In recent years,the importance of intelligent decision-making based on data is also increasing.The data-information-knowledge-wisdom is a new computing paradigm in intelligent decision-making based on big data.Data contain a wealth of information and create new possibilities for solving problems.Transforming data advantage into decision advantage as well as discovering its inherent law is one of the critical problems in intelligent data decision-making.The three-way decision model systematically studies the cognitive computing paradigm in computer science,decision science,management science,cognitive science,and other disciplines.The three-way decision is an information processing model that fits human cognition and an effective strategy to solve complex problems,which has important theoretical significance and practical value.This paper focuses on three-way decision model driven by data.We analyze the historical relationship and internal relationship between the three-way decision and granular computing.Aiming at specific practical problems,such as classification,attribute reduction,image recognition,there are many papers of three-way decision about these problems.However,it is still challenging to analyze the quantitative relationship between the three-way decision driven by data and the intelligent decision after discovering the observation space.On the one hand,it not conducive to the in-depth exploration of the inherent laws of data.On the other hand,it is not conducive to directly comparing the effects of different decision schemes.This paper is to study the three-way decision,including granular computing three-way decision,movement-based TAO model,change-based TAO model,and TAO model based on rough set by using the interval set,rough set,utility analysis,and other theoretical tools.The main innovations are as follows:(1)Research on data-driven TAO model and effectiveness measure method based on movement perspective,we propose the movement-based three-way decision of TAO model.Firstly,a medical diagnosis example is introduced to the three-way decision problems from the movement perspective,and the basic idea of solving complex problems by using the movement-based three-way decision is analyzed.The model includes the hierarchical structure of trisecting-acting-outcome,the movement strategy,the movement based on preference,and the movement process.Finally,a measurement method of the proportional utility function is proposed,in which the proportional value represents the ratio between the initial quantity and the final quantity.The movement-based measure method is discussed from coarse-grained to fine-grained.The simulation results and comparison results show the effectiveness and practicability of the method.(2)From the perspective of change,the TAO model of three-way decision in the cognitive background is studied,and we propose a change-based three-way decision model.Firstly,the basic components,explanations,and the relationship with the other threeway decision models are analyzed,and the TAO model of three-way decision based on change is studied.Then we discuss the construction forms of change model,including the change model based on interval set,the change model based on quantitative and qualitative change,and the change model based on evaluation function.Finally,a utility measure method for change-based three-way decision is established.The subjective cognition of decision-makers is considered in the effectiveness measure method,and it is transformed into quantifiable units for effect evaluation to form a utility measure framework for change models under the cognitive background.Example analysis and experimental results show that the model is effective and practical.(3)From the perspective of rough set to carry out the three-way decision model,research and application of TAO model are put forward,we propose change-based threeway decision model based on the rough set perspective.Based on the information system,we study the role of change-based three-way decision in the knowledge discovery method based on rough set.The row-column perspective in the information system is taken as the entry point to analyze and reason the system's data directly.Firstly,the application method of the changed-based three-way decision in rough set is discussed,and the representation method and semantic interpretation of the change model under qualitative and quantitative modes are studied,respectively.Then,from the perspective of the row,a classification strategy based on the change model is proposed.The classification rules and classification objects in the probabilistic approximate region are re-decided,and the three-way classification rules are reconstructed,and the rationality and effectiveness of the method are proved.Finally,from the perspectives of the column,we consider the properties of the change and apply it to the attribute reduction method,and put forward an attribute reduction method based on the change model.In the traditional global reduction conditions and integrated reduction separately under the condition of the experiment,the results show that the method under the condition of the two can protect classification accuracy and effectively reduce time consumption.