The Microsoft Linear Regression algorithm is a variant of the Microsoft Decision tree algorithm that helps you calculate the linear relationship between dependent and independent variables and then use that relationship for prediction.
The representation represented by the relationship is the formula that best represents the line of the data series. For example, the lines in the following diagram are the most likely linear representations of the data.
Each data point in the diagram has an error associated with the distance between the data point and the regression line. The coefficients A and b in the regression equation can adjust the angle and position of the regression line. You can adjust A and B until the total number of errors associated with all points reaches the lowest value to get the regression formula.
There are other types of linear regression that use multiple variables and nonlinear regression methods. However, linear regression is a well-known and useful way to model responses to changes in some of the underlying factors.
Example
You can use linear regression to determine the relationship between two contiguous columns. For example, you can use linear regression to calculate trend lines based on production or sales data. You can also use linear regression as a basis for developing more complex data mining models to evaluate the relationships between columns of data.
Although there are many ways to calculate linear regression, and these methods do not require data mining tools, the advantage of using the Microsoft linear regression algorithm to calculate linear regression is that you can automatically calculate and test all possible relationships between variables. You do not have to choose a calculation method, such as calculating the least squares method. However, for scenarios where the results are affected by multiple factors, linear regression may unduly simplify the relationship.
Here we go to the topic, through a simple process configuration we can implement the entire data mining process, followed by the following steps:
Reference documents:
Microsoft linear regression algorithm
http://msdn.microsoft.com/zh-cn/library/ms174824 (v=sql.105). aspx
"Bi thing" Microsoft linear regression algorithm