Forecasting Techniques
Below are forecasting techniques often used in Operations Research.
For series without trend and seasonality.
Model: where:
is independent and 
Moving Average

Where
|

|
= an estimate of M made at time t |
|
N |
= observations (given equal weight) |
|

|
= new observation |
Single Exponential Smoothing
Include all past data and give greater weight to more recent observations.


Where St = exponentially smoothed average at time t and

And,

|