Holt's's Two-Parameter Trend Model
Holt's's Two-Parameter Trend Model
1)
Compute level
St = α * Yt + (1 - α) * (St-1 + bt-1)
Where
α = Normal smoothed constant: alpha
St = smooted level in period t
2)
Compute trend
bt = γ * (St - St-1) + (1 - γ) * bt-1
Where
γ = Trend smoothed constant: gamma
bt = smooted trend in period t
3)
Forecast
Ft+p = St + p * bt
Where:
p = Number of periods forecasted (forecast horizon)