Autoregressive integrated moving average ARIMA Exponential smoothing;
12 often intertwined to generate new models. For example, the autoregressive
moving average model ARMA combines the AR model and the MA model. Another example of this is the autoregressive integrated moving average
ARIMA model, which combine all three of the models previously mentioned. The most commonly used model for time series data is the autoregressive process.
The autoregressive process is a difference equation determined by random variables. The distribution of such random variables is the key component in
modeling time series. Data collection techniques for discrete time series can be done in two ways, as
described as follows: 1. Through sampling of continuous time series. it means the continuous data is
sampled in the same time interval. 2. Through the accumulation of a variable in a period of time. For example,
rainfall is usually accumulated over a certain period of time days, months, etc..
Mathematical models for stochastic and deterministic dynamic problem: 1. If the value of a future future value of a time series can be appropriately
determined by a mathematical function, eg: Zt = cos 2πft, it can be called
deterministic time series. 2. If the future value can only be described in a probability distribution. It can be
called as a stochastic time series. The basic model used in the time series analysis is ARIMA that can be
expressed by a linear combination of observation variables data and independent random variables distributed normally. Based the describe above, the approaches
time series that uses to forecasting electricity demand in Libya are: