The data of general database is two-dimensional, while the general mathematical models and data analysis formula are also based on two-dimensional matrix. Nowadays, data sets based on data warehouse and some of the applied database (such as spatial database) are multi-dimensional, so there is an important problem of science and technology that is needed to be solved urgently, that is, to suit the expression of multi-dimensional matrix of multi-dimensional data sets, mathematical models and algorithms based on data warehouse. The stochastic frontier linear model, introduced and developed by statisticians, solves the objectivity and optimality problems of production function well. The main part is a linear model. In fact, linear regression may be largely deviated, and using single non-parameter regression alternatively may be inconvincible. Combined with linear regression and non-parameter regression, Stone firstly designed a model, which is popularly studied, named linear semi-parameter regression model. Now, we take advantage of this thought to study stochastic frontier problems, and present following stochastic frontier semi-parameter model. In the end of this paper, we give an example computed by DASC which is a software developed by ourself.
[1]
Chang Li,et al.
A data model for supporting on-line analytical processing
,
1996,
CIKM '96.
[2]
Jan Van den Bussche,et al.
A Graph-Oriented Object Database Model
,
1994,
IEEE Trans. Knowl. Data Eng..
[3]
Luca Cabibbo,et al.
Querying Multidimensional Databases
,
1997,
DBPL.
[4]
Sunita Sarawagi,et al.
Modeling multidimensional databases
,
1997,
Proceedings 13th International Conference on Data Engineering.
[5]
Anindya Datta,et al.
A Conceptual Model and Algebra for On-Line Analytical Processing in Decision Support Databases
,
2001,
Inf. Syst. Res..