17.7 Forecasting with Time Series Models

time series forecasting was made to predict beyond a point of time
huge field, focusing on ARMA here
AR: autoregression, MA: moveing average
AR(1): $Y_t = \beta_0 + \beta_1 Y_{t-1} + \epsilon$
MA(1): $Y_t = \beta_0 + (\epsilon_t + \theta\epsilon_{t-1})$
ARMA(2,1): $Y_t = \beta_0 + \beta_1 Y_{t-1} + \beta_1 Y_{t-2} + (\epsilon_t + \theta\epsilon_{t-1})$

{.unnumbered}