Dissertation > Social Sciences > Management > Decision Sciences

A Time-series Approach to Dynamic Multiple Criteria Decision Analysis with Application

Author ZhouMing
Tutor ChenZuo
School Nanjing University of Aeronautics and Astronautics
Course Systems Engineering
Keywords Multiple Criteria Decision Analysis Time series OWA(Ordered WeightedAveraging) Dynamic Multiple Criteria Decision Analysis The theory of evidence
CLC C934
Type Master's thesis
Year 2013
Downloads 12
Quotes 0
Download Dissertation

Decision-making as a strategy to achieve the intended goals, from all their options to find themost satisfactory method of decision-makers,is applied of the wide range.The practical problems of the complexity of the system need to be analyzed from various angles(attributes,criterias,targets,time),resulting in the lack of Multiple Criteria Decision in thedecision-making process.According to the reality, this article introduces the time factor preference, puts forwardMultiple-Criteria problems, systematically considering historical data, real state and thecomprehensive study of the future predicting results, and discusses the evidence to the theory ofdecision-making results.The main innovative points are as follows:(1)In the Multiple Criteria Decision Analysis problems, introduce the concept of time, to build aOWA-based dynamic Multiple Criteria Decision Analysis model, considering the decision resultsfrom the point of view of the three-dimensional space of the attributes, indicators and time;(2)Extends the decision-making of original historical data and the time point to thecomprehensive considering of the historical data, the point-in-time and future data, which has acertain sense of innovation, and has important theoretical guiding significance in the decision-makingof many complex issues of reality;(3)Introduce the future data to construct a predictive model based on accuracy optimization, andthe accuracy of the predicted results in the overall decision-making will be mapped to the final resultsof the comprehensive evaluation;(4)Integrate historical data, now nodes and future data to build the accuracy optimization modelof the evidential reasoning, given the confidence interval according to decision results, to get the bestpossible program.The problems are given numerical example, to verify the validity of the model, and to lay thefoundation for further research.

Related Dissertations
More Dissertations