Kostenloser Exponential Smoothing Rechner
Apply exponential smoothing to forecast time-series data. kostenloser Rechner mit adjustable alpha smoothing factor.
Next Forecast
113.0000
Next Forecast vs Latest Actual Value
Formel
## How to Apply Exponential Smoothing ### Formula **F(t+1) = alpha * A(t) + (1 - alpha) * F(t)** Equivalently: F(t+1) = F(t) + alpha * (A(t) - F(t)) Exponential smoothing updates the forecast by adding a fraction (alpha) of the latest forecast error. Alpha near 1 means the forecast reacts quickly to changes; alpha near 0 means it changes slowly. All past data contributes, with exponentially decreasing weights.
Lösungsbeispiel
Previous forecast = 110, actual = 120, alpha = 0.3.
- 01Error = 120 - 110 = 10
- 02Adjustment = 0.3 * 10 = 3
- 03New forecast = 110 + 3 = 113
- 04Or equivalently: 0.3 * 120 + 0.7 * 110 = 36 + 77 = 113
Häufig Gestellte Fragen
How do I choose the alpha value?
Higher alpha (0.7-0.9) gives more weight to recent data, suitable for volatile series. Lower alpha (0.1-0.3) smooths more aggressively, suitable for stable series. Optimize alpha by minimizing forecast error on historical data.
What is double exponential smoothing?
Double (Holt) exponential smoothing adds a second component for trend, allowing forecasts to follow upward or downward trends rather than always lagging behind. Triple (Holt-Winters) adds a seasonal component.
How does exponential smoothing compare to moving averages?
Both smooth data, but exponential smoothing gives more weight to recent observations and requires only the latest actual value and previous forecast (not the full window of data). It is more memory-efficient and often more responsive.
Lernen