Scribd
Upload
51 views24 pages

Smoothing

The document summarizes smoothing methods for time series forecasting. It discusses naive models, average methods like simple and moving averages, exponential smoothing methods including single, double, and triple exponential smoothing, and the Winter's method. These methods are applied to real data examples in Minitab to illustrate how they can be implemented and evaluate forecast accuracy using error metrics. The document provides an overview of key smoothing techniques for forecasting time series data.

Uploaded by

nani
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
51 views24 pages

Smoothing

The document summarizes smoothing methods for time series forecasting. It discusses naive models, average methods like simple and moving averages, exponential smoothing methods including single, double, and triple exponential smoothing, and the Winter's method. These methods are applied to real data examples in Minitab to illustrate how they can be implemented and evaluate forecast accuracy using error metrics. The document provides an overview of key smoothing techniques for forecasting time series data.

Uploaded by

nani
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 24

Metode Pemulusan (Smoothing)

Pengajar:Gumgum Darmawan,M.Si

Metode Pemulusan merupakan metode


peramalan yang didasarkan pada pola data
secara empiris untuk meramalkan masa yang
akan datang.
Ciri khas dari metode pemulusan adalah data
yang baru mempunyai bobot yang lebih
besar dari data sebelumnya.
Pembobotan dari data membentuk pola
eksponensial.
Kaitan Pola Data dengan Metode Peramalan

Time Series Patterns

Trend and
Stationer Trend Effect Seasonal Effect
Seasonal

 Naïve Model  Naïve Model


 Simple  Double
 Naïve Model
Averages  Moving Averages

Moving Averages  Double
Winter’s Model
 Single Exponential Exponential Smoothing
Smoothing
Naïve Model

 The recent periods are the best predictors of the future.

1. The simplest model for stationary data is


Yˆt 1  Yt
2. The simplest model for trend data is
Yˆt 1  Yt  (Yt  Yt 1 ) or
ˆ Yt
Yt 1  Yt
Yt 1
3. The simplest model for seasonal data is
Yˆt 1  Y(t 1) s
MINITAB implementation

Time Series Plot


MINITAB implementation … (continued)

Naïve 1 Naïve 2 Naïve 3


MINITAB implementation … (continued)

Naïve 1

Naïve 2

Naïve 3
MINITAB implementation … (continued)

Naïve 1

Naïve 2

Naïve 3

MSE.1 = 28547.5, MSE.2 = 53592.5, MSE.3 = 4567.5


Measuring Forecasting Error …

 MSE (mean squared error)


 rata-rata kuadrat kesalahan (residual atau error).
n
MSE   ei2 / n
i 1
 MAD (mean absolute deviation)
 ukuran kesalahan peramalan dalam unit ukuran yang
sama dengan data aslinya.

n
MAD   ei / n
i 1
Measuring Forecasting Error …

 MAPE (mean absolute percentage error)


 persentase kesalahan absolut rata-rata.

n
MAPE   PEi / n
i 1

 MPE (mean percentage error)


 persentase kesalahan rata-rata.

n
 ei 
MPE   PEi / n PEi     100 
i 1  Xi 
Average Methods

1. Simple Averages
 obtained by finding the mean for all the relevant values
and then using this mean to forecast the next period.
n Yt
Yt 1  
ˆ for stationary data
t 1 n

2. Moving Averages
 obtained by finding the mean for a specified set of
values and then using this mean to forecast
the next period.
(Y  Y    Yt  n 1 )
St  Yˆt 1  t t 1 for stationary data
n
Moving Average
Average Methods … (continued)

3. Linear Moving Averages(LMA)


 one set of moving averages is computed, and then a second
set is computed as a moving average of the first set.
(Y  Y    Yt  n 1 )
(i). St  Yˆt 1  t t 1
n
( St  St 1    St  n 1 )
S 
(ii). t 
n

(iii).at  2 St  St
2
(iv).bt  ( St  St )
N 1

Ft  m  at  bt m for a linear trend data


Aplikasi Linear Moving Average
Eksponensial Tunggal:
Pendekatan Adaptif
Pemulusan Eksponensial tunggal dengan tingkat respon yang
adptif (ARRSES) memiliki kelebihan dibandingkan Single
Eksponensial Tunggal (SES) yaitu nilai α yang berubah secara
otomatis jika terdapat perubahan pada pola data.

Parameter α dan β bernilai antara 0 dan 1.


Aplikasi Metode ARRSES
Exponential Smoothing Methods

 Single Exponential Smoothing  for stationary data


Ft 1   X t  (1   ) Ft

 Exponential Smoothing Adjusted for Trend : Holt’s Method


1. The exponentially smoothed series :
St =  Xt + (1) (St-1+ bt-1)

2. The trend estimate :


bt   ( S t  St 1 )  (1   )bt 1
3. Forecast p periods into the future :
Ft  m  St  mbt
Aplikasi Metode Holt
Pemulusan Eksponensial Ganda: Metode
Linier satu parameter dari Brown

Metode Brown satu parameter memperbaiki metode


LMA jika terdapat unsur trend.
Aplikasi Metode Linier Satu
Parameter dari Brown
Pemulusan Eksponensial Tripel:
Metode Kuadratik Satu-Parameter dari Brown

Digunakan untuk meramalkan data dengan


pola trend kuadratik, kubik atau orde yang
lebih tinggi.
Aplikasi Metode Pemulusan Tripel: Metode
Kuadratik Satu parameter dari Brown.
Exponential Smoothing Adjusted for Trend and
Seasonal Variation : Winter’s Method

1. The exponentially smoothed series :


Xt
St    (1   ) ( St 1  bt 1 )
It  L
2. The trend estimate :
Three
bt   ( S t  St 1 )  (1   )bt 1 parameters
models
3. The seasonality estimate :
Xt
It    (1   ) I t 1
St

4. Forecast p periods into the future :


Ft  m  ( St  mbt ) It  L  m
Aplikasi Metode Winters
Wassalam

The End of This Session

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy