Arima¶

• for data shows clear trends or seasonal patterns

• ideal for short-term forecasts

• can use differencing to remove non-stationary trend

SARIMAX example¶

https://github.com/SamWachira/Electricity-Demand-Forecasting/blob/main/Time_Series_Electricity.ipynb

ARIMAX (autoregressive integrated moving average with exogenous variables)¶

• The ARIMAX model is an extension of the standard ARIMA model, and

• it allows for the inclusion of one or more exogenous variables in the model equation.

• The ARIMAX model can be estimated using maximum likelihood estimation, and

• the parameters of the model can be estimated using numerical optimization techniques.

components¶

• Auto-Regressive (AR):
• determine the present value based on historical values

• only for data with stationary trend

• Moving Average (MA):

• smoothen variations to reveal the underlying trend

• only for data with stationary trend

• Integrated (I):

• integrate the AR and MA elements by computing differences between current and past values
• transform the dataset into a stationary trend

parameters¶

• p: the number of past values considered for the AR component

• q: the number of moving averages applied

• d: the number of past values subject to differentiation

limit¶

lacking adaptability