Making forecasting easier forecast v7 for Outline 1 Motivation - - PowerPoint PPT Presentation

▶
making forecasting easier forecast v7 for outline
SMART_READER_LITE
LIVE PREVIEW

Making forecasting easier forecast v7 for Outline 1 Motivation - - PowerPoint PPT Presentation

Apr 15, 2023 162 likes •753 views

Rob J Hyndman Making forecasting easier forecast v7 for Outline 1 Motivation and history 2 Automatic forecasting in R 3 ggplot2 graphics 4 Bias adjustment Making forecasting easier Motivation and history 2 Motivation Making

slide-1
SLIDE 1

Rob J Hyndman

Making forecasting easier forecast v7 for

slide-2
SLIDE 2

Outline

1 Motivation and history 2 Automatic forecasting in R 3 ggplot2 graphics 4 Bias adjustment

Making forecasting easier Motivation and history 2

slide-3
SLIDE 3

Motivation

Making forecasting easier Motivation and history 3

slide-4
SLIDE 4

Motivation

Making forecasting easier Motivation and history 3

slide-5
SLIDE 5

Motivation

Making forecasting easier Motivation and history 3

slide-6
SLIDE 6

Motivation

Making forecasting easier Motivation and history 3

slide-7
SLIDE 7

Motivation

Making forecasting easier Motivation and history 3

slide-8
SLIDE 8

Motivation

1

Common in business to have thousands of products that need forecasting at least monthly.

2

Forecasts are often required by people who are untrained in time series analysis. Specifications Automatic forecasting algorithms must:

➥ determine an appropriate time series model; ➥ estimate the parameters; ➥ compute the forecasts with prediction intervals.

Making forecasting easier Motivation and history 4

slide-9
SLIDE 9

Motivation

1

Common in business to have thousands of products that need forecasting at least monthly.

2

Forecasts are often required by people who are untrained in time series analysis. Specifications Automatic forecasting algorithms must:

➥ determine an appropriate time series model; ➥ estimate the parameters; ➥ compute the forecasts with prediction intervals.

Making forecasting easier Motivation and history 4

slide-10
SLIDE 10

Forecast package history

Pre 2003 Collection of functions used for consulting projects July/August 2003 ets and thetaf added August 2006 v1.0 available on CRAN May 2007 auto.arima added July 2008 JSS paper (Hyndman & Khandakar) September 2009 v2.0. Unbundled. May 2010 arfima added Feb/March 2011 tslm, stlf, naive, snaive added August 2011 v3.0. Box Cox transformations added December 2011 tbats added April 2012 Package moved to github November 2012 v4.0. nnetar added June 2013 Major speed-up of ets January 2014 v5.0. tsoutliers and tsclean added May 2015 v6.0. Added several new plots December 2015 264,000 package downloads in one month! February 2016 v7.0. Added ggplot2 graphics & bias adjustment

Making forecasting easier Motivation and history 5

slide-11
SLIDE 11

Outline

1 Motivation and history 2 Automatic forecasting in R 3 ggplot2 graphics 4 Bias adjustment

Making forecasting easier Automatic forecasting in R 6

slide-12
SLIDE 12

Automatic methods in forecast package

Making forecasting easier Automatic forecasting in R 7

Automatic model selection auto.arima + forecast ets + forecast tbats + forecast bats + forecast arfima + forecast ar + forecast nnetar + forecast stlm + forecast Automatic forecasting forecast.ts stlf thetaf dshw, hw, holt, ses splinef rwf, naive croston All produce an object of class “forecast”

slide-13
SLIDE 13

ets algorithm in R

Based on Hyndman, Koehler, Snyder & Grose (IJF 2002): For each model, optimize parameters and initial values of underlying state space model using MLE. Select best method using AICc. Produce forecasts and prediction intervals using best method.

Making forecasting easier Automatic forecasting in R 8

slide-14
SLIDE 14

ets algorithm in R

Based on Hyndman, Koehler, Snyder & Grose (IJF 2002): For each model, optimize parameters and initial values of underlying state space model using MLE. Select best method using AICc. Produce forecasts and prediction intervals using best method.

Making forecasting easier Automatic forecasting in R 8

slide-15
SLIDE 15

ets algorithm in R

Based on Hyndman, Koehler, Snyder & Grose (IJF 2002): For each model, optimize parameters and initial values of underlying state space model using MLE. Select best method using AICc. Produce forecasts and prediction intervals using best method.

Making forecasting easier Automatic forecasting in R 8

slide-16
SLIDE 16

Exponential smoothing

Making forecasting easier Automatic forecasting in R 9

Forecasts from ETS(M,A,N)

1960 1970 1980 1990 2000 2010 300 400 500 600

slide-17
SLIDE 17

Exponential smoothing

fit <- ets(livestock) fcast <- forecast(fit) plot(fcast)

Making forecasting easier Automatic forecasting in R 10

Forecasts from ETS(M,A,N)

1960 1970 1980 1990 2000 2010 300 400 500 600

slide-18
SLIDE 18

Exponential smoothing

Making forecasting easier Automatic forecasting in R 11

Forecasts from ETS(M,Ad,M)

1995 2000 2005 2010 0.4 0.6 0.8 1.0 1.2 1.4

slide-19
SLIDE 19

Exponential smoothing

fit <- ets(h02) fcast <- forecast(fit) plot(fcast)

Making forecasting easier Automatic forecasting in R 12

Forecasts from ETS(M,Ad,M)

1995 2000 2005 2010 0.4 0.6 0.8 1.0 1.2 1.4

slide-20
SLIDE 20

auto.arima algorithm in R

Based on Hyndman and Khandakar (JSS 2008): Select no. differences via unit root tests. Use stepwise search to traverse model space, starting with a simple model and considering nearby variants. For each model, optimize parameters using MLE. Select best method using AICc. Produce forecasts and prediction intervals using best method.

Making forecasting easier Automatic forecasting in R 13

slide-21
SLIDE 21

auto.arima algorithm in R

Based on Hyndman and Khandakar (JSS 2008): Select no. differences via unit root tests. Use stepwise search to traverse model space, starting with a simple model and considering nearby variants. For each model, optimize parameters using MLE. Select best method using AICc. Produce forecasts and prediction intervals using best method.

Making forecasting easier Automatic forecasting in R 13

slide-22
SLIDE 22

auto.arima algorithm in R

Based on Hyndman and Khandakar (JSS 2008): Select no. differences via unit root tests. Use stepwise search to traverse model space, starting with a simple model and considering nearby variants. For each model, optimize parameters using MLE. Select best method using AICc. Produce forecasts and prediction intervals using best method.

Making forecasting easier Automatic forecasting in R 13

slide-23
SLIDE 23

auto.arima algorithm in R

Based on Hyndman and Khandakar (JSS 2008): Select no. differences via unit root tests. Use stepwise search to traverse model space, starting with a simple model and considering nearby variants. For each model, optimize parameters using MLE. Select best method using AICc. Produce forecasts and prediction intervals using best method.

Making forecasting easier Automatic forecasting in R 13

slide-24
SLIDE 24

auto.arima algorithm in R

Based on Hyndman and Khandakar (JSS 2008): Select no. differences via unit root tests. Use stepwise search to traverse model space, starting with a simple model and considering nearby variants. For each model, optimize parameters using MLE. Select best method using AICc. Produce forecasts and prediction intervals using best method.

Making forecasting easier Automatic forecasting in R 13

slide-25
SLIDE 25

Auto ARIMA

Making forecasting easier Automatic forecasting in R 14

Forecasts from ARIMA(0,1,0) with drift

1960 1970 1980 1990 2000 2010 250 300 350 400 450 500 550

slide-26
SLIDE 26

Auto ARIMA

fit <- auto.arima(livestock) fcast <- forecast(fit) plot(fcast)

Making forecasting easier Automatic forecasting in R 15

Forecasts from ARIMA(0,1,0) with drift

1960 1970 1980 1990 2000 2010 250 300 350 400 450 500 550

slide-27
SLIDE 27

Auto ARIMA

Making forecasting easier Automatic forecasting in R 16

Forecasts from ARIMA(3,1,3)(0,1,1)[12]

1995 2000 2005 2010 0.4 0.6 0.8 1.0 1.2 1.4

slide-28
SLIDE 28

Auto ARIMA

fit <- auto.arima(h02) fcast <- forecast(fit) plot(fcast)

Making forecasting easier Automatic forecasting in R 17

Forecasts from ARIMA(3,1,3)(0,1,1)[12]

1995 2000 2005 2010 0.4 0.6 0.8 1.0 1.2 1.4

slide-29
SLIDE 29

TBATS model TBATS Trigonometric terms for seasonality Box-Cox transformations for heterogeneity ARMA errors for short-term dynamics Trend (possibly damped) Seasonal (including multiple and non-integer periods)

Automatic algorithm described in De Livera, Hyndman and Snyder (JASA 2011).

Making forecasting easier Automatic forecasting in R 18

slide-30
SLIDE 30

Examples

fit <- tbats(gas) fcast <- forecast(fit) plot(fcast)

Making forecasting easier Automatic forecasting in R 19

Forecasts from TBATS(1, {0,0}, 1, {<52.1785714285714,9>})

1995 2000 2005 7 8 9 10

slide-31
SLIDE 31

Examples

fit <- tbats(callcentre) fcast <- forecast(fit) plot(fcast)

Making forecasting easier Automatic forecasting in R 20

Forecasts from TBATS(0.607, {0,0}, −, {<169,5>, <845,4>})

2005 2010 2015 2020 2025 2030 2035 100 200 300 400 500

slide-32
SLIDE 32

Outline

1 Motivation and history 2 Automatic forecasting in R 3 ggplot2 graphics 4 Bias adjustment

Making forecasting easier ggplot2 graphics 21

slide-33
SLIDE 33

ggplot2 graphics

Making forecasting easier ggplot2 graphics 22

Internet Usage per Minute

Minutes Internet users 20 40 60 80 100 100 150 200

plot(WWWusage, xlab="Minutes", ylab="Internet users", main="Internet Usage per Minute")

slide-34
SLIDE 34

ggplot2 graphics

Making forecasting easier ggplot2 graphics 23

80 120 160 200 25 50 75 100

Minutes Internet users

Internet Usage per Minute

autoplot(WWWusage, xlab="Minutes", ylab="Internet users", main="Internet Usage per Minute")

slide-35
SLIDE 35

ggplot2 graphics

Making forecasting easier ggplot2 graphics 24

100 150 200 250 300 30 60 90

Time y level

80 95

Forecasts from ETS(A,Ad,N)

autoplot(forecast(WWWusage))

slide-36
SLIDE 36

ggplot2 graphics

Making forecasting easier ggplot2 graphics 25

100 150 200 250 300 30 60 90

Time y level

80 95

Forecasts from ETS(A,Ad,N)

library(magrittr) WWWusage %>% forecast %>% autoplot

slide-37
SLIDE 37

ggplot2 graphics

Making forecasting easier ggplot2 graphics 26

100 150 200 250 30 60 90

Minute Number of users

Forecasts from ARIMA model

WWWusage %>% auto.arima %>% forecast(level=c(50,80,95)) -> fc autoplot(WWWusage) + geom_forecast(fc, color=’#ffcccc’, show.legend=FALSE) + ggtitle("Forecasts from ARIMA model") + labs(x="Minute", y="Number of users")

slide-38
SLIDE 38

ggplot2 graphics

autoplot methods ts forecast acf stl Arima ets . . . Other ggplot2 graphics ggseasonplot ggmonthplot ggtsdisplay

Making forecasting easier ggplot2 graphics 27

slide-39
SLIDE 39

Outline

1 Motivation and history 2 Automatic forecasting in R 3 ggplot2 graphics 4 Bias adjustment

Making forecasting easier Bias adjustment 28

slide-40
SLIDE 40

Box-Cox transformations

wt = fλ(yt) =

  • log(yt),

λ = 0; (yλ

t − 1)/λ,

λ = 0. λ = 1: (No substantive transformation) λ = 1

2: (Square root plus linear transformation)

λ = 0: (Natural logarithm) λ = −1: (Inverse plus 1)

Most modelling and forecasting functions in the forecast package have a lambda argument allowing Box-Cox transformations.

Making forecasting easier Bias adjustment 29

slide-41
SLIDE 41

Box-Cox transformations

wt = fλ(yt) =

  • log(yt),

λ = 0; (yλ

t − 1)/λ,

λ = 0. λ = 1: (No substantive transformation) λ = 1

2: (Square root plus linear transformation)

λ = 0: (Natural logarithm) λ = −1: (Inverse plus 1)

Most modelling and forecasting functions in the forecast package have a lambda argument allowing Box-Cox transformations.

Making forecasting easier Bias adjustment 29

slide-42
SLIDE 42

Box-Cox transformations

wt = fλ(yt) =

  • log(yt),

λ = 0; (yλ

t − 1)/λ,

λ = 0. λ = 1: (No substantive transformation) λ = 1

2: (Square root plus linear transformation)

λ = 0: (Natural logarithm) λ = −1: (Inverse plus 1)

Most modelling and forecasting functions in the forecast package have a lambda argument allowing Box-Cox transformations.

Making forecasting easier Bias adjustment 29

slide-43
SLIDE 43

Box-Cox transformations

wt = fλ(yt) =

  • log(yt),

λ = 0; (yλ

t − 1)/λ,

λ = 0. λ = 1: (No substantive transformation) λ = 1

2: (Square root plus linear transformation)

λ = 0: (Natural logarithm) λ = −1: (Inverse plus 1)

Most modelling and forecasting functions in the forecast package have a lambda argument allowing Box-Cox transformations.

Making forecasting easier Bias adjustment 29

slide-44
SLIDE 44

Box-Cox transformations

wt = fλ(yt) =

  • log(yt),

λ = 0; (yλ

t − 1)/λ,

λ = 0. λ = 1: (No substantive transformation) λ = 1

2: (Square root plus linear transformation)

λ = 0: (Natural logarithm) λ = −1: (Inverse plus 1)

Most modelling and forecasting functions in the forecast package have a lambda argument allowing Box-Cox transformations.

Making forecasting easier Bias adjustment 29

slide-45
SLIDE 45

Box-Cox transformations

wt = fλ(yt) =

  • log(yt),

λ = 0; (yλ

t − 1)/λ,

λ = 0. λ = 1: (No substantive transformation) λ = 1

2: (Square root plus linear transformation)

λ = 0: (Natural logarithm) λ = −1: (Inverse plus 1)

Most modelling and forecasting functions in the forecast package have a lambda argument allowing Box-Cox transformations.

Making forecasting easier Bias adjustment 29

slide-46
SLIDE 46

Box-Cox transformations

wt = fλ(yt) =

  • log(yt),

λ = 0; (yλ

t − 1)/λ,

λ = 0. λ = 1: (No substantive transformation) λ = 1

2: (Square root plus linear transformation)

λ = 0: (Natural logarithm) λ = −1: (Inverse plus 1)

Most modelling and forecasting functions in the forecast package have a lambda argument allowing Box-Cox transformations.

Making forecasting easier Bias adjustment 29

slide-47
SLIDE 47

Box-Cox transformations

Making forecasting easier Bias adjustment 30

λ = 1.00

Year Monthly electricity production 1960 1970 1980 1990 2000 6000 10000 14000

slide-48
SLIDE 48

Back-transformation

We must reverse the transformation (or back-transform) to obtain forecasts on the original

  • scale. The reverse Box-Cox transformations are

given by yt = f −1

λ (wt) =

  • exp(wt),

λ = 0; (λwt + 1)1/λ, λ = 0.

fit <- auto.arima(elec, lambda=1/3) fc <- forecast(fit) plot(fc, include=120)

Making forecasting easier Bias adjustment 31

slide-49
SLIDE 49

Back-transformation

Making forecasting easier Bias adjustment 32

Forecasts from ARIMA(0,1,1)(0,1,1)[12]

1986 1988 1990 1992 1994 1996 1998 10000 12000 14000 16000

slide-50
SLIDE 50

Bias and Box-Cox

If the forecast is the mean on the transformed scale, it is not the mean on the original scale. If the forecast is the median on the transformed scale, it is the median on the original scale. Quantiles are preserved because the transformation is monotonically increasing. If E(W) = µ and Var(W) = σ2, then Bias E(Y)−eµ ≈

  • (λµ + 1)1/λ

1 + σ2(1−λ)

2(λµ+1)2

  • − eµ

if λ = 0;

1 2eµσ2

if λ = 0.

Making forecasting easier Bias adjustment 33

slide-51
SLIDE 51

Bias and Box-Cox

If the forecast is the mean on the transformed scale, it is not the mean on the original scale. If the forecast is the median on the transformed scale, it is the median on the original scale. Quantiles are preserved because the transformation is monotonically increasing. If E(W) = µ and Var(W) = σ2, then Bias E(Y)−eµ ≈

  • (λµ + 1)1/λ

1 + σ2(1−λ)

2(λµ+1)2

  • − eµ

if λ = 0;

1 2eµσ2

if λ = 0.

Making forecasting easier Bias adjustment 33

slide-52
SLIDE 52

Bias and Box-Cox

If the forecast is the mean on the transformed scale, it is not the mean on the original scale. If the forecast is the median on the transformed scale, it is the median on the original scale. Quantiles are preserved because the transformation is monotonically increasing. If E(W) = µ and Var(W) = σ2, then Bias E(Y)−eµ ≈

  • (λµ + 1)1/λ

1 + σ2(1−λ)

2(λµ+1)2

  • − eµ

if λ = 0;

1 2eµσ2

if λ = 0.

Making forecasting easier Bias adjustment 33

slide-53
SLIDE 53

Bias and Box-Cox

If the forecast is the mean on the transformed scale, it is not the mean on the original scale. If the forecast is the median on the transformed scale, it is the median on the original scale. Quantiles are preserved because the transformation is monotonically increasing. If E(W) = µ and Var(W) = σ2, then Bias E(Y)−eµ ≈

  • (λµ + 1)1/λ

1 + σ2(1−λ)

2(λµ+1)2

  • − eµ

if λ = 0;

1 2eµσ2

if λ = 0.

Making forecasting easier Bias adjustment 33

slide-54
SLIDE 54

Bias and Box-Cox

If the forecast is the mean on the transformed scale, it is not the mean on the original scale. If the forecast is the median on the transformed scale, it is the median on the original scale. Quantiles are preserved because the transformation is monotonically increasing. If E(W) = µ and Var(W) = σ2, then Bias E(Y)−eµ ≈

  • (λµ + 1)1/λ

1 + σ2(1−λ)

2(λµ+1)2

  • − eµ

if λ = 0;

1 2eµσ2

if λ = 0.

Making forecasting easier Bias adjustment 33

slide-55
SLIDE 55

Bias adjustment

Making forecasting easier Bias adjustment 34

100 200 300 1900 1950 2000 2050

Time cents

Price of a dozen eggs

fit <- ets(eggs, lambda=0) fc <- forecast(fit, h=50, level=95) fc2 <- forecast(fit, h=50, level=95, biasadj=TRUE) autoplot(fc, main="Price of a dozen eggs", ylab="cents") + geom_forecast(fc2, plot.conf=FALSE, color="red") + guides(fill=FALSE)

slide-56
SLIDE 56

Bias adjustment

Making forecasting easier Bias adjustment 35

100 200 300 1900 1950 2000 2050

Time cents

Price of a dozen eggs

slide-57
SLIDE 57

For further information

robjhyndman.com

Slides for this talk. Links to all papers and books. Links to R packages. A blog about forecasting research.

OTexts.org/fpp

Free online book based on forecast package for R.

Making forecasting easier Bias adjustment 36