|
|
人们在实验和生产过程中，会得到很多二维以及二维以上的相关数据，
这些数据反过来帮助人们解决实际中的问题，需要进行数据处理，使之成为反映这些数据变化规律的数学摸型。 应用“最小二乘法”回归数据只能做线性回归，对于非线性问题，要通过对过程假设，建立相关性质数学关系式，即机理模型，对机理模型进行线性化处理，再做回归建模,计算机理模型中各元的系数。回归后的模型，有些数据相关性很好，但实际中的数据千变万化，有些推导出机理模型，线性化处理后，回归的模型相关性也不好，甚至有些相关数据，根本就推导不出机理模型，回归建立数学模型就更困难了。
“最小三乘法”
解决了“最小二乘法”在回归相关数据中的问题。
依据“最小三乘法”
开发的数据回归建模软件 “DRS”，使得一元线性、多元线性、一元非线性以至多元非线性的数据回归(三维的曲面类数据和更多维的复杂数据回归），计算更简单结果更准确。 |
|
People will obtain many
relating data of two or more than two dimension during
experiments and production. These data will help
them to solve problems of reality on contrary,
which need data processing to make them
become mathematicalematical model
reflecting the data
variation regulation.
The application of the "Least
Square Method" can only make linear regression, but
to the nonlinear problems it must construct relating
mathematicalematical relationship expression, namely mechanism
model through procedure supposing to do linearization processing
of mechanism model and then do regression modeling computation.
Some relating data of the recursive models are good, but the
data of reality are changeable, some deduce mechanism models.
After the linear process the correlation property of the
regression model is not good, and some relating data even can't
deduce in the mechanism model. It is even more harder to build
mathematical models.
"Least
Cubic Method" solves problems that "Least
Square Method" Data Regression met in the regression of relating data. Based on "Least Cubic Method" the development
of data regression modeling software "DRS", which
made the computation and results for data regression of one
variable linear, multivariate linear and one variable nonlinear
and multivariate nonlinear data calculating more easy and
correct.
|
