Predictive Analytics – Exponential Smoothing for Forecasting
Predictive Analytics
Exponential Smoothing for Forecasting
Copyright By PowCoder代写 加微信 powcoder
Discipline of Business Analytics, The University of School
This week: Exponential Smoothing
1. Simple exponential smoothing
2. Trend corrected exponential smoothing
3. Holt winters smoothing
4. Damped trend exponential smoothing
Chapter 8 of the book: https://otexts.com/fpp3/expsmooth.html
Exponential smoothing methods
Exponential smoothing forecasts are weighted averages of past
observations, where the weights decay exponentially as we go
further into the past.
Exponential smoothing can be useful when the time series
components are changing over time.
Simple exponential smoothing
Simple exponential smoothing
The simple exponential smoothing (SES) method specifies the
forecasting rule
ŷt+1 = ℓt (forecast equation)
ℓt = αyt + (1 − α)ℓt−1 (smoothing equation)
for an initial value ℓ0 and 0 ≤ α ≤ 1.
ℓt is known as the level of the time series.
Exponentially weighted moving average
ℓ1 = αy1 + (1 − α)ℓ0
ℓ2 = αy2 + (1 − α)ℓ1
= αy2 + (1 − α)αy1 + (1 − α)2ℓ0
ℓ3 = αy3 + (1 − α)ℓ2
= αy3 + (1 − α)αy2 + (1 − α)2αy1 + (1 − α)3ℓ0
ℓ4 = αy4 + (1 − α)ℓ3
= αy4 + (1 − α)αy3 + (1 − α)2αy2 + (1 − α)3αy1 + (1 − α)4ℓ0
Exponentially weighted moving average
It follows that
ℓt =αyt + (1 − α)ℓt−1
=αyt + (1 − α)αyt−1 + (1 − α)2αyt−2 + . . . + (1 − α)t−1αy1
+ (1 − α)tℓ0.
Simple exponential smoothing is also known as the exponentially
weighted moving average (EWMA) method.
Simple exponential smoothing
• Useful for forecasting time series with changing levels.
• A higher α gives larger weight to recent observations, making
the forecasts more adaptive to recent changes in the series.
• A lower α leads to a larger weights for past observations,
making the forecasts smoother.
• Initialisation: we typically set ℓ0 = y1 for simplicity.
Alternatively, we can treat it as a parameter.
Example: AUD/USD exchange rate
Example: AUD/USD exchange rate
Estimation
We estimate α by least squares (empirical risk minimisation).
α̂ = argmin
(yt − ℓt−1)2
Each ℓt is a nonlinear function of α, so that there is no formula for
α̂. We use numerical optimisation methods to obtain the solution.
Statistical model
In order to say more about the simple exponential smoothing
method, we need to formulate it as a statistical model. We assume
Yt = ℓt−1 + εt,
ℓt = αyt + (1 − α)ℓt−1,
where the errors εt are i.i.d with constant variance σ2.
Statistical model
In forecasting, we want to:
1. compute point forecasts for multiple forecasting horizons h.
2. compute interval forecasts for multiple forecasting horizons h.
In order to do this for the exponential smoothing method, we
rewrite the model in error correction form.
Error correction form
We obtain the error correction form as
ℓt = αYt + (1 − α)ℓt−1
= ℓt−1 + α(Yt − ℓt−1)
= ℓt−1 + αεt.
Hence, we can rewrite the model as:
Yt+1 = ℓt + εt+1,
ℓt = ℓt−1 + αεt.
Error correction form
Using ℓt = ℓt−1 + αεt,
ℓt+1 = ℓt + αεt+1
ℓt+2 = ℓt+1 + αεt+2
= ℓt + αεt+1 + αεt+2
ℓt+3 = ℓt+2 + αεt+3
= ℓt + αεt+1 + αεt+2 + αεt+3
ℓt+h = ℓt +
Constant plus noise representation
Using Yt = ℓt−1 + εt and the previous slide,
Yt+1 = ℓt + εt+1
Yt+2 = ℓt+1 + εt+2
= ℓt + αεt+1 + εt+2
Yt+3 = ℓt+2 + εt+3
= ℓt + αεt+1 + αεt+2 + εt+3
Yt+h = ℓt+h−1 + εt+h
αεt+i + εt+h
Point forecast
Constant plus noise representation of future observations:
Yt+h = ℓt +
αεt+i + εt+h
From the linearity of expectations, the point forecast for any
horizon h is
ŷt+h = E(Yt+h|y1:t)
αεt+i + εt+h
∣∣∣∣∣ y1:t
Forecast variance
Var(Yt+1|y1:t) = Var(ℓt + εt+1|y1:t)
Var(Yt+2|y1:t) = Var(ℓt + αεt+1 + εt+2|y1:t)
= σ2(1 + α2)
Var(Yt+h|y1:t) = Var
αεt+h−i + εt+h
∣∣∣∣∣ y1:t
= σ2(1 + (h − 1)α2)
Forecast equations for simple exponential smoothing
ŷt+h = ℓt
Var(Yt+h|y1:t) = σ2(1 + (h − 1)α2)
Interval forecast
If we assume that εt ∼ N(0, σ2),
Yt+h|y1:t ∼ N
1 + (h − 1)α2
To compute an interval forecast, we use the estimated values of α
ℓ̂t ± zcrit ×
σ̂2 [1 + (h − 1)α̂2],
t=1(yt − ℓt−1)2
If the errors are not normal, you should use the Bootstrap method
or other distributional assumptions.
Example: AUD/USD exchange rate
Trend corrected exponential
Trend corrected exponential smoothing
The trend corrected or Holt exponential smoothing method
allows for a time-varying trend:
ŷt+1 = ℓt + bt (forecast equation)
ℓt = αyt + (1 − α)(ℓt−1 + bt−1) (smoothing equation)
bt = β(ℓt − ℓt−1) + (1 − β)bt−1 (trend equation)
for an initial values ℓ0 and b0, 0 ≤ α ≤ 1, and 0 ≤ β ≤ 1.
Trend corrected exponential smoothing
Consider the simple time series trend model
ℓt = a + b × t,
Yt = ℓt + εt.
What is ℓt − ℓt−1 here?
Trend corrected exponential smoothing model
The statistical model is
Yt+1 = ℓt + bt + εt+1,
ℓt = αYt + (1 − α)(ℓt−1 + bt−1),
bt = β(ℓt − ℓt−1) + (1 − β)bt−1,
where the errors εt are i.i.d with constant variance σ2.
The least squares estimates of α and β are
α̂, β̂ = argmin
(yt − ℓt−1 − bt−1)2
Error correction form
ℓt = αYt + (1 − α)(ℓt−1 + bt−1)
= ℓt−1 + bt−1 + α(Yt − ℓt−1 − bt−1)
= ℓt−1 + bt−1 + αεt
bt = β(ℓt − ℓt−1) + (1 − β)bt−1
= bt−1 + β(ℓt − ℓt−1 − bt−1)
= bt−1 + βα(ℓt−1 + bt−1 + αεt − ℓt−1 − bt−1)
= bt−1 + βαεt
Error correction form
Yt+1 = ℓt + bt + εt+1
ℓt = ℓt−1 + bt−1 + αεt
bt = bt−1 + βαεt
Constant plus noise representation
Yt+1 = ℓt + bt + εt+1
Yt+2 = ℓt+1 + bt+1 + εt+2
= ℓt + 2bt + α(1 + β)εt+1 + εt+2
Yt+3 = ℓt+2 + bt+2 + +εt+3
= ℓt+1 + 2bt+1 + α(1 + β)εt+2 + εt+3
= ℓt + 3bt + α(1 + 2β)εt+1 + α(1 + β)εt+2 + εt+3
Yt+h = ℓt + hbt + α
(1 + iβ)εt+h−i + εt+h
Point forecast
Constant plus noise representation of future observations:
Yt+h = ℓt + hbt + α
(1 + iβ)εt+h−i + εt+h
From the linearity of expectations, the point forecast for any
horizon h is
ŷt+h = E(Yt+h|y1:t)
ℓt + hbt + α
(1 + iβ)εt+h−i + εt+h
∣∣∣∣∣ y1:t
= ℓt + hbt.
Forecast variance
Var(Yt+1|y1:t) = Var(ℓt + bt + εt+1|y1:t)
Var(Yt+2|y1:t) = Var(ℓt + 2bt + α(1 + β)εt+1 + εt+2|y1:t)
= σ2(1 + α2(1 + β)2)
Var(Yt+h|y1:t) = Var
ℓt + hbt + α
(1 + iβ)εt+h−i + εt+h|y1:t
Forecast equations for the trend corrected smoothing method
Point forecast:
ŷt+h = ℓ̂t + hb̂t
Var(Yt+h|y1:t) = σ2
We compute interval forecasts as before.
Example: assaults in Sydney
Example: assaults in Sydney
Example: visitor arrivals
Example: visitor arrivals
• Exponential smoothing methods for forecast
• Simple exponential smoothing
• Trend corrected exponential smoothing
• Error correction forms for constructing forecasts: point
forecast, interval forecast and density forecast.
Holt winters smoothing
exponential smoothing
The Holt-Winters exponential smoothing method extend the
trend corrected method to seasonal data. It allows for additive or
multiplicative seasonality.
Additive Smoothing (key concept)
ŷt+1 = ℓt + bt + St+1−L (forecast equation)
ℓt = α(yt − St−L) + (1 − α)(ℓt−1 + bt−1) (level)
bt = β(ℓt − ℓt−1) + (1 − β)bt−1, (trend)
St = δ(yt − ℓt) + (1 − δ)St−L, (seasonal indices)
for a seasonal frequency L, initial values ℓ0, b0, and Si−L for
i = 1, . . . , L, and parameters 0 ≤ α ≤ 1, 0 ≤ β ≤ 1, 0 ≤ δ ≤ 1.
Multiplicative Smoothing (key concept)
ŷt+1 = (ℓt + bt) × St+1−L (forecast equation)
ℓt = α(yt/St−L) + (1 − α)(ℓt−1 + bt−1) (level)
bt = β(ℓt − ℓt−1) + (1 − β)bt−1, (trend)
St = δ(yt/ℓt) + (1 − δ)St−L, (seasonal indices)
for a seasonal frequency L, initial values ℓ0, b0, and Si−L for
i = 1, . . . , L, and parameters 0 ≤ α ≤ 1, 0 ≤ β ≤ 1, 0 ≤ δ ≤ 1.
Additive v.s. Multiplicative
• Additive smoothing is
appropriate when the
variation (around the trend,
or the seasonal pattern)
doesn’t change over time
• Multiplicative smoothing is
appropriate when the
variation (around the trend,
or the seasonal pattern)
changes over time.
Statistical model
As before, we formulate a statistical model by specifying an
observation equation.
Yt+1 = ℓt + bt + St+1−L + εt+1,
where εt+1 is i.i.d with variance σ2
Multiplicative:
yt+1 = (ℓt + bt) × St+1−L + εt+1,
where εt+1 is i.i.d with variance σ2.
Estimation
We estimate α, β and δ by least squares.
α̂, β̂, δ̂ = argmin
(yt − ℓt−1 − bt−1 − St−L)2
Multiplicative:
α̂, β̂, δ̂ = argmin
(yt − (ℓt−1 + bt−1) × St−L)2
Forecast equations
ŷt+h = ℓ̂t + hb̂t + St−L+(h mod L)
Var(Yt+h|y1:t) = σ2
[α(1 + iβ) + Ii,Lδ(1 − α)]2
where mod is the modulo operator, Ii,L = 0 if h mod L ̸= i and
Ii,L = 1 if h mod L = i.
Multiplicative:
ŷt+h = (ℓ̂t + hb̂t) × St−L+(h mod L)
No simple expression exists for the variance in the multiplicative
Example: assaults in estimated parameters are α̂ = 0.117, β̂ = 0.023, and
δ̂ = 0.370.
Example: visitor arrivals
The estimated parameters are α̂ = 0.154, β̂ = 0.088, and
δ̂ = 0.271.
Example: visitor arrivals
Damped trend exponential
Damped trend exponential smoothing
Damped trend exponential smoothing addresses the problem
that extrapolating trends indefinitely into the future can lead to
implausible forecasts.
Model and forecast
yt+1 = ℓt + ϕbt + εt+1,
ℓt = αyt + (1 − α)(ℓt−1 + ϕbt−1),
bt = β(ℓt − ℓt−1) + (1 − β)ϕbt−1,
where ϕ is the damping parameter, with 0 ≤ ϕ ≤ 1.
Forecast equation:
ŷt+h = ℓt + ϕbt + ϕ2bt + ϕ3bt + . . . + ϕhbt
We can extend it to allow for additive or multiplicative seasonality.
Illustration: visitor arrivals
Review questions
• What is exponential smoothing?
• What is the difference between simple, trend corrected, and
Holt-Winters exponential smoothing methods?
• Derive the point forecasts and forecast variances for the SES
and trend corrected methods, starting from the model
equations.
• Explain how to compute forecast intervals based on the SES
and trend corrected methods.
Simple exponential smoothing
Trend corrected exponential smoothing
Holt winters smoothing
Damped trend exponential smoothing
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com