APPLIED RESEARCH OF INFORMATION IN DEPOSIT FORECAST
Because a mass of correlations appear among various economic variables in the practical economic system, it is possible to adopt data fusion technology in deposit forecast and analysis.The paper develops an elementary attempt on the application research of information fusion in the deposit forecast Aiming at the problems existing in the current statistical forecast methods for economic variables, especially the extrapolation forecast method based on polynomial fitting technology, the basic framework for deposit forecast based on data fusion is established by introducing data fusion technology and some modeling ways.Compared with the method based on the own history of target variable and polynomial fitting technology, the proposed method has many advantages such as higher forecast accuracy and better stability and reliability.Also, the simple computer simulation based on actual deposit data from Nantong City of China shows the validity of the proposed forecast method.But, some important issues should be further researched, for example much more precise solution for the parameters and the expression for correlations among these variables.
Information fusion Deposit forecast Correlations Modeling Kalman fitler
ZHEN-YU ZHU XI-ZHAO ZHOU
Economy and Administration College, Shanghai Maritime University, Shanghai, 200135, China
国际会议
2007 International Conference on Machine Learning and Cybernetics(IEEE第六届机器学习与控制论国际会议)
香港
英文
1784-1789
2007-08-19（万方平台首次上网日期，不代表论文的发表时间）