difference equation that formulates the discrete autoregressive and moving average 
process ARMA(2.1). 
The above mentioned is for the simple case of ARMA(2.1) versus AR(2). For the 
more complicated cases of ARMA(4.3), ARMA(6.5) ..., theoretically A(4), 
A(6),..., Should correspond, and the parameters of differential equations that formu- 
late these A(4), A(6)... should be derived from the parameters of the difference 
equations that formulate ARMA(4.3), ARMA(6.5) ..., although their relations might 
be much more complicated than for A(2) from ARMA(2.1). The differential equations 
that formulate the continuous autoregressive process give us good clues to finding the 
physical characteristics of their response processes. 
The second order linear differential equation is the basic equation of a linear dynam- 
ic system with one degree of freedom. Accordingly, the above discussion indicates that 
the ARMA(2.1) process represents the dynamic behavior of a linear system with one 
degree of freedom under the excitement of white noise. 
226 
