The grey model can be used to predict effectively, especially when the data is blurred, such as predicting when to rain, first to explain a simple and most commonly used GM model GM (a)
Generate columns
Set the original time series to X1 (t) T for time t=0,1,2,3 ...., generating the column x0 (t) is the difference that the original sequence of the t>=1 is:
x0(t)=x1(t)-x1(t-1)
Immediate Value column
As the name implies, the average value is taken.
z1=0.5*x10.5*x1(t-1)
GM (model)
In fact, the generated column is the derivative of the original sequence:
x1(t)‘=x0(t) =x1(t)-x1(t-1)
We can build relationships as follows:
A, B is constant
Order B=[-Z1, 1]
U=[a, b]. T
a*z1=b a*z1=bx0=B*uu=(B.T*B).I*B.T*x0x1(k+1)=(x0(1)-b/a)*exp(-a*k)+a/b
So we can make predictions.
Copyright NOTICE: This article for Bo Master original article, without Bo Master permission not reproduced.
Grey prediction of model GM model in Mathematical modeling model (I.)