Forecasting is a process of showing future trends by identifying regular patterns existing data values.
The technique of identifying regular patterns from existing data values is known as exponential smoothing . Tableau uses exponential smoothing and gives an accurate forecast graph that they can use in predictive analysis.
Forecasting in Tableau is based on two important concepts i.e. Trends and Seasonality .
A trend is the increase or decrease in data over time and seasonality is a repeating variation in values over a determined or fixed period of time (such as weekly, quarterly, yearly, etc). These are known as seasons.
These seasonal and repeating variations are predictable giving us accurate forecasts for predictive analysis.