DocumentCode :
1797958
Title :
Impact of data to decision based on emergency plan
Author :
Shengli Pan ; Jun Tian
Author_Institution :
Inf. Manage. & E-Bus. Dept., Xi´an Jiaotong Univ., Xi´an, China
fYear :
2014
fDate :
15-17 Nov. 2014
Firstpage :
979
Lastpage :
986
Abstract :
This paper studies the sequential decision under constraint conditions. This paper studies relation of status data, data quality to decision results, data worth and the rules of the designated solution. This paper uses simulation model, the Bayesian formula and the minimal expectations value principle, analyzes characteristics of data to decision results by changing the value of different model parameters. This paper´s differences are constructing sequential decision under constraint conditions, and introducing match variable instead of the single variable, replacing expectations maximum principle with expectations minimum principle. This paper´s findings show that high quality and stable solutions depends on the change of which state data and(or) data quality play the leading role. This phenomenon is interpreted through the closure property of status data value and data quality value, and the expected value principle. The merit of this paper lies in that we not only extract and solve one category important but unsighted decision problem, but also construct a set formula to analyze the relation of data to decision quality. Our conslusions are universal to the above decision category and not limited to emergency decision.
Keywords :
Bayes methods; data handling; decision making; emergency management; expectation-maximisation algorithm; Bayesian formula; data quality value; decision quality; emergency decision; emergency plan; expectation maximum principle; expectation minimum principle; minimal expectation value principle; sequential decision; status data value; Accuracy; Analytical models; Bayes methods; Data models; Loss measurement; Stability analysis; Supply and demand; Bayesian decision; data; decision quality; match variables; the minimal expected value principle;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Systems and Informatics (ICSAI), 2014 2nd International Conference on
Conference_Location :
Shanghai
Print_ISBN :
978-1-4799-5457-5
Type :
conf
DOI :
10.1109/ICSAI.2014.7009428
Filename :
7009428
Link To Document :
بازگشت