8. Motivation
1
2
Common in business to have over 1000
products that need forecasting at least monthly.
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.
Automatic time series forecasting
Motivation
4
9. Motivation
1
2
Common in business to have over 1000
products that need forecasting at least monthly.
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.
Automatic time series forecasting
Motivation
4
14. Outline
1 Motivation
2 Forecasting competitions
3 Forecasting the PBS
4 Time series forecasting
5 Evaluating forecast accuracy
6 Exponential smoothing
7 ARIMA modelling
8 Automatic nonlinear forecasting?
9 Time series with complex seasonality
10 Forecasts about automatic forecasting
Automatic time series forecasting
Forecasting competitions
7
15. Makridakis and Hibon (1979)
Automatic time series forecasting
Forecasting competitions
8
16. Makridakis and Hibon (1979)
Automatic time series forecasting
Forecasting competitions
8
17. Makridakis and Hibon (1979)
This was the first large-scale empirical evaluation of
time series forecasting methods.
Highly controversial at the time.
Difficulties:
How to measure forecast accuracy?
How to apply methods consistently and objectively?
How to explain unexpected results?
Common thinking was that the more
sophisticated mathematical models (ARIMA
models at the time) were necessarily better.
If results showed ARIMA models not best, it
must be because analyst was unskilled.
Automatic time series forecasting
Forecasting competitions
9
18. Makridakis and Hibon (1979)
I do not believe that it is very fruitful to attempt to
classify series according to which forecasting techniques
perform “best”. The performance of any particular
technique when applied to a particular series depends
essentially on (a) the model which the series obeys;
(b) our ability to identify and fit this model correctly and
(c) the criterion chosen to measure the forecasting
accuracy.
— M.B. Priestley
. . . the paper suggests the application of normal scientific
experimental design to forecasting, with measures of
unbiased testing of forecasts against subsequent reality,
for success or failure. A long overdue reform.
— F.H. Hansford-Miller
Automatic time series forecasting
Forecasting competitions
10
19. Makridakis and Hibon (1979)
I do not believe that it is very fruitful to attempt to
classify series according to which forecasting techniques
perform “best”. The performance of any particular
technique when applied to a particular series depends
essentially on (a) the model which the series obeys;
(b) our ability to identify and fit this model correctly and
(c) the criterion chosen to measure the forecasting
accuracy.
— M.B. Priestley
. . . the paper suggests the application of normal scientific
experimental design to forecasting, with measures of
unbiased testing of forecasts against subsequent reality,
for success or failure. A long overdue reform.
— F.H. Hansford-Miller
Automatic time series forecasting
Forecasting competitions
10
20. Makridakis and Hibon (1979)
Modern man is fascinated with the subject of
forecasting — W.G. Gilchrist
It is amazing to me, however, that after all this
exercise in identifying models, transforming and so
on, that the autoregressive moving averages come
out so badly. I wonder whether it might be partly
due to the authors not using the backwards
forecasting approach to obtain the initial errors.
— W.G. Gilchrist
Automatic time series forecasting
Forecasting competitions
11
21. Makridakis and Hibon (1979)
Modern man is fascinated with the subject of
forecasting — W.G. Gilchrist
It is amazing to me, however, that after all this
exercise in identifying models, transforming and so
on, that the autoregressive moving averages come
out so badly. I wonder whether it might be partly
due to the authors not using the backwards
forecasting approach to obtain the initial errors.
— W.G. Gilchrist
Automatic time series forecasting
Forecasting competitions
11
22. Makridakis and Hibon (1979)
I find it hard to believe that Box-Jenkins, if properly
applied, can actually be worse than so many of the
simple methods — C. Chatfield
Why do empirical studies sometimes give different
answers? It may depend on the selected sample of
time series, but I suspect it is more likely to depend
on the skill of the analyst and on their individual
interpretations of what is meant by Method X.
— C. Chatfield
. . . these authors are more at home with simple
procedures than with Box-Jenkins. — C. Chatfield
Automatic time series forecasting
Forecasting competitions
12
23. Makridakis and Hibon (1979)
I find it hard to believe that Box-Jenkins, if properly
applied, can actually be worse than so many of the
simple methods — C. Chatfield
Why do empirical studies sometimes give different
answers? It may depend on the selected sample of
time series, but I suspect it is more likely to depend
on the skill of the analyst and on their individual
interpretations of what is meant by Method X.
— C. Chatfield
. . . these authors are more at home with simple
procedures than with Box-Jenkins. — C. Chatfield
Automatic time series forecasting
Forecasting competitions
12
24. Makridakis and Hibon (1979)
I find it hard to believe that Box-Jenkins, if properly
applied, can actually be worse than so many of the
simple methods — C. Chatfield
Why do empirical studies sometimes give different
answers? It may depend on the selected sample of
time series, but I suspect it is more likely to depend
on the skill of the analyst and on their individual
interpretations of what is meant by Method X.
— C. Chatfield
. . . these authors are more at home with simple
procedures than with Box-Jenkins. — C. Chatfield
Automatic time series forecasting
Forecasting competitions
12
25. Makridakis and Hibon (1979)
There is a fact that Professor Priestley must accept:
empirical evidence is in disagreement with his
theoretical arguments. — S. Makridakis M. Hibon
Dr Chatfield expresses some personal views about
the first author . . . It might be useful for Dr Chatfield
to read some of the psychological literature quoted
in the main paper, and he can then learn a little
more about biases and how they affect prior
probabilities.
— S. Makridakis M. Hibon
Automatic time series forecasting
Forecasting competitions
13
26. Makridakis and Hibon (1979)
There is a fact that Professor Priestley must accept:
empirical evidence is in disagreement with his
theoretical arguments. — S. Makridakis M. Hibon
Dr Chatfield expresses some personal views about
the first author . . . It might be useful for Dr Chatfield
to read some of the psychological literature quoted
in the main paper, and he can then learn a little
more about biases and how they affect prior
probabilities.
— S. Makridakis M. Hibon
Automatic time series forecasting
Forecasting competitions
13
27. Consequences of MH (1979)
As a result of this paper, researchers started to:
¯
¯
¯
¯
consider how to automate forecasting methods;
study what methods give the best forecasts;
be aware of the dangers of over-fitting;
treat forecasting as a different problem from
time series analysis.
Makridakis Hibon followed up with a new
competition in 1982:
1001 series
Anyone could submit forecasts (avoiding the
charge of incompetence)
Multiple forecast measures used.
Automatic time series forecasting
Forecasting competitions
14
28. Consequences of MH (1979)
As a result of this paper, researchers started to:
¯
¯
¯
¯
consider how to automate forecasting methods;
study what methods give the best forecasts;
be aware of the dangers of over-fitting;
treat forecasting as a different problem from
time series analysis.
Makridakis Hibon followed up with a new
competition in 1982:
1001 series
Anyone could submit forecasts (avoiding the
charge of incompetence)
Multiple forecast measures used.
Automatic time series forecasting
Forecasting competitions
14
30. M-competition
Main findings (taken from Makridakis Hibon, 2000)
1
Statistically sophisticated or complex methods do
not necessarily provide more accurate forecasts
than simpler ones.
2
The relative ranking of the performance of the
various methods varies according to the accuracy
measure being used.
3
The accuracy when various methods are being
combined outperforms, on average, the individual
methods being combined and does very well in
comparison to other methods.
4
The accuracy of the various methods depends upon
the length of the forecasting horizon involved.
Automatic time series forecasting
Forecasting competitions
16
32. Makridakis and Hibon (2000)
“The M3-Competition is a final attempt by the authors to
settle the accuracy issue of various time series methods. . .
The extension involves the inclusion of more methods/
researchers (in particular in the areas of neural networks
and expert systems) and more series.”
3003 series
All data from business, demography, finance and
economics.
Series length between 14 and 126.
Either non-seasonal, monthly or quarterly.
All time series positive.
MH claimed that the M3-competition supported the
findings of their earlier work.
However, best performing methods far from “simple”.
Automatic time series forecasting
Forecasting competitions
18
33. Makridakis and Hibon (2000)
Best methods:
Theta
A very confusing explanation.
Shown by Hyndman and Billah (2003) to be average of
linear regression and simple exponential smoothing
with drift, applied to seasonally adjusted data.
Later, the original authors claimed that their
explanation was incorrect.
Forecast Pro
A commercial software package with an unknown
algorithm.
Known to fit either exponential smoothing or ARIMA
models using BIC.
Automatic time series forecasting
Forecasting competitions
19
34. M3 results (recalculated)
Method
MAPE sMAPE MASE
Theta
ForecastPro
ForecastX
Automatic ANN
B-J automatic
17.42
18.00
17.35
17.18
19.13
Automatic time series forecasting
12.76
13.06
13.09
13.98
13.72
Forecasting competitions
1.39
1.47
1.42
1.53
1.54
20
35. M3 results (recalculated)
Method
MAPE sMAPE MASE
Theta
17.42 12.76
ForecastPro
18.00 13.06
ForecastX
17.35 13.09
® Calculations 17.18 13.98
Automatic ANN do not match
published
B-J automatic paper.
19.13 13.72
1.39
1.47
1.42
1.53
1.54
® Some contestants apparently
submitted multiple entries but only
best ones published.
Automatic time series forecasting
Forecasting competitions
20
36. Outline
1 Motivation
2 Forecasting competitions
3 Forecasting the PBS
4 Time series forecasting
5 Evaluating forecast accuracy
6 Exponential smoothing
7 ARIMA modelling
8 Automatic nonlinear forecasting?
9 Time series with complex seasonality
10 Forecasts about automatic forecasting
Automatic time series forecasting
Forecasting the PBS
21
38. Forecasting the PBS
The Pharmaceutical Benefits Scheme (PBS) is
the Australian government drugs subsidy scheme.
Many drugs bought from pharmacies are
subsidised to allow more equitable access to
modern drugs.
The cost to government is determined by the
number and types of drugs purchased.
Currently nearly 1% of GDP ($14 billion).
The total cost is budgeted based on forecasts
of drug usage.
Automatic time series forecasting
Forecasting the PBS
23
39. Forecasting the PBS
The Pharmaceutical Benefits Scheme (PBS) is
the Australian government drugs subsidy scheme.
Many drugs bought from pharmacies are
subsidised to allow more equitable access to
modern drugs.
The cost to government is determined by the
number and types of drugs purchased.
Currently nearly 1% of GDP ($14 billion).
The total cost is budgeted based on forecasts
of drug usage.
Automatic time series forecasting
Forecasting the PBS
23
40. Forecasting the PBS
The Pharmaceutical Benefits Scheme (PBS) is
the Australian government drugs subsidy scheme.
Many drugs bought from pharmacies are
subsidised to allow more equitable access to
modern drugs.
The cost to government is determined by the
number and types of drugs purchased.
Currently nearly 1% of GDP ($14 billion).
The total cost is budgeted based on forecasts
of drug usage.
Automatic time series forecasting
Forecasting the PBS
23
41. Forecasting the PBS
The Pharmaceutical Benefits Scheme (PBS) is
the Australian government drugs subsidy scheme.
Many drugs bought from pharmacies are
subsidised to allow more equitable access to
modern drugs.
The cost to government is determined by the
number and types of drugs purchased.
Currently nearly 1% of GDP ($14 billion).
The total cost is budgeted based on forecasts
of drug usage.
Automatic time series forecasting
Forecasting the PBS
23
42. Forecasting the PBS
In 2001: $4.5 billion budget, under-forecasted
by $800 million.
Thousands of products. Seasonal demand.
Subject to covert marketing, volatile products,
uncontrollable expenditure.
Although monthly data available for 10 years,
data are aggregated to annual values, and only
the first three years are used in estimating the
forecasts.
All forecasts being done with the FORECAST
function in MS-Excel!
Automatic time series forecasting
Forecasting the PBS
24
43. Forecasting the PBS
In 2001: $4.5 billion budget, under-forecasted
by $800 million.
Thousands of products. Seasonal demand.
Subject to covert marketing, volatile products,
uncontrollable expenditure.
Although monthly data available for 10 years,
data are aggregated to annual values, and only
the first three years are used in estimating the
forecasts.
All forecasts being done with the FORECAST
function in MS-Excel!
Automatic time series forecasting
Forecasting the PBS
24
44. Forecasting the PBS
In 2001: $4.5 billion budget, under-forecasted
by $800 million.
Thousands of products. Seasonal demand.
Subject to covert marketing, volatile products,
uncontrollable expenditure.
Although monthly data available for 10 years,
data are aggregated to annual values, and only
the first three years are used in estimating the
forecasts.
All forecasts being done with the FORECAST
function in MS-Excel!
Automatic time series forecasting
Forecasting the PBS
24
45. Forecasting the PBS
In 2001: $4.5 billion budget, under-forecasted
by $800 million.
Thousands of products. Seasonal demand.
Subject to covert marketing, volatile products,
uncontrollable expenditure.
Although monthly data available for 10 years,
data are aggregated to annual values, and only
the first three years are used in estimating the
forecasts.
All forecasts being done with the FORECAST
function in MS-Excel!
Automatic time series forecasting
Forecasting the PBS
24
46. Forecasting the PBS
In 2001: $4.5 billion budget, under-forecasted
by $800 million.
Thousands of products. Seasonal demand.
Subject to covert marketing, volatile products,
uncontrollable expenditure.
Although monthly data available for 10 years,
data are aggregated to annual values, and only
the first three years are used in estimating the
forecasts.
All forecasts being done with the FORECAST
function in MS-Excel!
Automatic time series forecasting
Forecasting the PBS
24
47. Total cost: A03 concession safety net group
600
400
200
0
$ thousands
800
1000
1200
PBS data
1995
2000
2005
Time
Automatic time series forecasting
Forecasting the PBS
25
49. Total cost: D01 general copayments group
400
300
200
100
0
$ thousands
500
600
700
PBS data
1995
2000
2005
Time
Automatic time series forecasting
Forecasting the PBS
25
52. Outline
1 Motivation
2 Forecasting competitions
3 Forecasting the PBS
4 Time series forecasting
5 Evaluating forecast accuracy
6 Exponential smoothing
7 ARIMA modelling
8 Automatic nonlinear forecasting?
9 Time series with complex seasonality
10 Forecasts about automatic forecasting
Automatic time series forecasting
Time series forecasting
26
55. Time series forecasting
Inputs
Output
y t −1
y t −2
yt
y t −3
Autoregression moving
average (ARMA) model
εt
ε t −1
ε t −2
Automatic time series forecasting
Time series forecasting
27
56. Time series forecasting
Inputs
Output
y t −1
y t −2
yt
y t −3
εt
ε t −1
ε t −2
Autoregression moving
average (ARMA) model
Estimation
Compute likelihood L from
ε1 , ε2 , . . . , εT .
Use optimization
algorithm to maximize L.
Automatic time series forecasting
Time series forecasting
27
57. Time series forecasting
xt −1
yt
εt
Automatic time series forecasting
State space model
xt is unobserved.
Time series forecasting
28
58. Time series forecasting
xt −1
yt
εt
xt
Automatic time series forecasting
State space model
xt is unobserved.
Time series forecasting
28
59. Time series forecasting
xt −1
yt
εt
xt
yt+1
State space model
xt is unobserved.
εt+1
Automatic time series forecasting
Time series forecasting
28
60. Time series forecasting
xt −1
yt
εt
xt
yt+1
εt+1
xt+1
Automatic time series forecasting
State space model
xt is unobserved.
Time series forecasting
28
61. Time series forecasting
xt −1
yt
εt
xt
yt+1
State space model
xt is unobserved.
εt+1
xt+1
yt+2
εt+2
Automatic time series forecasting
Time series forecasting
28
62. Time series forecasting
xt −1
yt
εt
xt
yt+1
State space model
xt is unobserved.
εt+1
xt+1
yt+2
εt+2
xt +2
Automatic time series forecasting
Time series forecasting
28
63. Time series forecasting
xt −1
yt
εt
xt
yt+1
State space model
xt is unobserved.
εt+1
xt+1
yt+2
εt+2
xt +2
y t +3
εt+3
Automatic time series forecasting
Time series forecasting
28
64. Time series forecasting
xt −1
yt
εt
xt
yt+1
State space model
xt is unobserved.
εt+1
xt+1
yt+2
εt+2
xt +2
y t +3
εt+3
xt +3
Automatic time series forecasting
Time series forecasting
28
65. Time series forecasting
xt −1
yt
εt
xt
yt+1
State space model
xt is unobserved.
εt+1
xt+1
yt+2
εt+2
xt +2
y t +3
εt+3
xt +3
y t +4
ε t +4
Automatic time series forecasting
Time series forecasting
28
66. Time series forecasting
xt −1
yt
εt
xt
yt+1
State space model
xt is unobserved.
εt+1
xt+1
yt+2
εt+2
xt +2
y t +3
εt+3
xt +3
Estimation
Compute likelihood L from
ε1 , ε2 , . . . , εT .
Use optimization
algorithm to maximize L.
Automatic time series forecasting
y t +4
ε t +4
Time series forecasting
28
67. Outline
1 Motivation
2 Forecasting competitions
3 Forecasting the PBS
4 Time series forecasting
5 Evaluating forecast accuracy
6 Exponential smoothing
7 ARIMA modelling
8 Automatic nonlinear forecasting?
9 Time series with complex seasonality
10 Forecasts about automatic forecasting
Automatic time series forecasting
Evaluating forecast accuracy
29
73. Akaike’s Information Criterion
AIC = −2 log(L) + 2k
where L is the likelihood and k is the number of
estimated parameters in the model.
This is a penalized likelihood approach.
If L is Gaussian, then AIC ≈ c + T log MSE + 2k
where c is a constant, MSE is from one-step
forecasts on training set, and T is the length of
the series.
Minimizing the Gaussian AIC is asymptotically
equivalent (as T → ∞) to minimizing MSE from
one-step forecasts on test set via time series
cross-validation.
Automatic time series forecasting
Evaluating forecast accuracy
31
74. Akaike’s Information Criterion
AIC = −2 log(L) + 2k
where L is the likelihood and k is the number of
estimated parameters in the model.
This is a penalized likelihood approach.
If L is Gaussian, then AIC ≈ c + T log MSE + 2k
where c is a constant, MSE is from one-step
forecasts on training set, and T is the length of
the series.
Minimizing the Gaussian AIC is asymptotically
equivalent (as T → ∞) to minimizing MSE from
one-step forecasts on test set via time series
cross-validation.
Automatic time series forecasting
Evaluating forecast accuracy
31
75. Akaike’s Information Criterion
AIC = −2 log(L) + 2k
where L is the likelihood and k is the number of
estimated parameters in the model.
This is a penalized likelihood approach.
If L is Gaussian, then AIC ≈ c + T log MSE + 2k
where c is a constant, MSE is from one-step
forecasts on training set, and T is the length of
the series.
Minimizing the Gaussian AIC is asymptotically
equivalent (as T → ∞) to minimizing MSE from
one-step forecasts on test set via time series
cross-validation.
Automatic time series forecasting
Evaluating forecast accuracy
31
76. Akaike’s Information Criterion
AIC = −2 log(L) + 2k
where L is the likelihood and k is the number of
estimated parameters in the model.
This is a penalized likelihood approach.
If L is Gaussian, then AIC ≈ c + T log MSE + 2k
where c is a constant, MSE is from one-step
forecasts on training set, and T is the length of
the series.
Minimizing the Gaussian AIC is asymptotically
equivalent (as T → ∞) to minimizing MSE from
one-step forecasts on test set via time series
cross-validation.
Automatic time series forecasting
Evaluating forecast accuracy
31
77. Akaike’s Information Criterion
AIC = −2 log(L) + 2k
where L is the likelihood and k is the number of
estimated parameters in the model.
This is a penalized likelihood approach.
If L is Gaussian, then AIC ≈ c + T log MSE + 2k
where c is a constant, MSE is from one-step
forecasts on training set, and T is the length of
the series.
Minimizing the Gaussian AIC is asymptotically
equivalent (as T → ∞) to minimizing MSE from
one-step forecasts on test set via time series
cross-validation.
Automatic time series forecasting
Evaluating forecast accuracy
31
78. Akaike’s Information Criterion
AIC = −2 log(L) + 2k
Corrected AIC
For small T, AIC tends to over-fit. Bias-corrected
version:
AICC = AIC +
2(k +1)(k +2)
T −k
Bayesian Information Criterion
BIC = AIC + k [log(T ) − 2]
BIC penalizes terms more heavily than AIC
Minimizing BIC is consistent if there is a true
model.
Automatic time series forecasting
Evaluating forecast accuracy
32
79. Akaike’s Information Criterion
AIC = −2 log(L) + 2k
Corrected AIC
For small T, AIC tends to over-fit. Bias-corrected
version:
AICC = AIC +
2(k +1)(k +2)
T −k
Bayesian Information Criterion
BIC = AIC + k [log(T ) − 2]
BIC penalizes terms more heavily than AIC
Minimizing BIC is consistent if there is a true
model.
Automatic time series forecasting
Evaluating forecast accuracy
32
80. Akaike’s Information Criterion
AIC = −2 log(L) + 2k
Corrected AIC
For small T, AIC tends to over-fit. Bias-corrected
version:
AICC = AIC +
2(k +1)(k +2)
T −k
Bayesian Information Criterion
BIC = AIC + k [log(T ) − 2]
BIC penalizes terms more heavily than AIC
Minimizing BIC is consistent if there is a true
model.
Automatic time series forecasting
Evaluating forecast accuracy
32
81. What to use?
Choice: AIC, AICc, BIC, CV-MSE
CV-MSE too time consuming for most automatic
forecasting purposes. Also requires large T.
As T → ∞, BIC selects true model if there is
one. But that is never true!
AICc focuses on forecasting performance, can
be used on small samples and is very fast to
compute.
Empirical studies in forecasting show AIC is
better than BIC for forecast accuracy.
Automatic time series forecasting
Evaluating forecast accuracy
33
82. What to use?
Choice: AIC, AICc, BIC, CV-MSE
CV-MSE too time consuming for most automatic
forecasting purposes. Also requires large T.
As T → ∞, BIC selects true model if there is
one. But that is never true!
AICc focuses on forecasting performance, can
be used on small samples and is very fast to
compute.
Empirical studies in forecasting show AIC is
better than BIC for forecast accuracy.
Automatic time series forecasting
Evaluating forecast accuracy
33
83. What to use?
Choice: AIC, AICc, BIC, CV-MSE
CV-MSE too time consuming for most automatic
forecasting purposes. Also requires large T.
As T → ∞, BIC selects true model if there is
one. But that is never true!
AICc focuses on forecasting performance, can
be used on small samples and is very fast to
compute.
Empirical studies in forecasting show AIC is
better than BIC for forecast accuracy.
Automatic time series forecasting
Evaluating forecast accuracy
33
84. What to use?
Choice: AIC, AICc, BIC, CV-MSE
CV-MSE too time consuming for most automatic
forecasting purposes. Also requires large T.
As T → ∞, BIC selects true model if there is
one. But that is never true!
AICc focuses on forecasting performance, can
be used on small samples and is very fast to
compute.
Empirical studies in forecasting show AIC is
better than BIC for forecast accuracy.
Automatic time series forecasting
Evaluating forecast accuracy
33
85. What to use?
Choice: AIC, AICc, BIC, CV-MSE
CV-MSE too time consuming for most automatic
forecasting purposes. Also requires large T.
As T → ∞, BIC selects true model if there is
one. But that is never true!
AICc focuses on forecasting performance, can
be used on small samples and is very fast to
compute.
Empirical studies in forecasting show AIC is
better than BIC for forecast accuracy.
Automatic time series forecasting
Evaluating forecast accuracy
33
86. Outline
1 Motivation
2 Forecasting competitions
3 Forecasting the PBS
4 Time series forecasting
5 Evaluating forecast accuracy
6 Exponential smoothing
7 ARIMA modelling
8 Automatic nonlinear forecasting?
9 Time series with complex seasonality
10 Forecasts about automatic forecasting
Automatic time series forecasting
Exponential smoothing
34
89. Exponential smoothing methods
Trend
Component
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
Automatic time series forecasting
Exponential smoothing
36
90. Exponential smoothing methods
Trend
Component
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
N,N:
Simple exponential smoothing
Automatic time series forecasting
Exponential smoothing
36
91. Exponential smoothing methods
Trend
Component
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
N,N:
A,N:
Simple exponential smoothing
Holt’s linear method
Automatic time series forecasting
Exponential smoothing
36
92. Exponential smoothing methods
Trend
Component
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
N,N:
A,N:
Ad ,N:
Simple exponential smoothing
Holt’s linear method
Additive damped trend method
Automatic time series forecasting
Exponential smoothing
36
93. Exponential smoothing methods
Trend
Component
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
N,N:
A,N:
Ad ,N:
M,N:
Simple exponential smoothing
Holt’s linear method
Additive damped trend method
Exponential trend method
Automatic time series forecasting
Exponential smoothing
36
94. Exponential smoothing methods
Trend
Component
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
N,N:
A,N:
Ad ,N:
M,N:
Md ,N:
Simple exponential smoothing
Holt’s linear method
Additive damped trend method
Exponential trend method
Multiplicative damped trend method
Automatic time series forecasting
Exponential smoothing
36
95. Exponential smoothing methods
Trend
Component
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
N,N:
A,N:
Ad ,N:
M,N:
Md ,N:
A,A:
Simple exponential smoothing
Holt’s linear method
Additive damped trend method
Exponential trend method
Multiplicative damped trend method
Additive Holt-Winters’ method
Automatic time series forecasting
Exponential smoothing
36
96. Exponential smoothing methods
Trend
Component
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
N,N:
A,N:
Ad ,N:
M,N:
Md ,N:
A,A:
A,M:
Simple exponential smoothing
Holt’s linear method
Additive damped trend method
Exponential trend method
Multiplicative damped trend method
Additive Holt-Winters’ method
Multiplicative Holt-Winters’ method
Automatic time series forecasting
Exponential smoothing
36
97. Exponential smoothing methods
Trend
Component
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
There are 15 separate exponential smoothing
methods.
Automatic time series forecasting
Exponential smoothing
36
98. Exponential smoothing methods
Trend
Component
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
There are 15 separate exponential smoothing
methods.
Each can have an additive or multiplicative error,
giving 30 separate models.
Automatic time series forecasting
Exponential smoothing
36
99. Exponential smoothing methods
Trend
Component
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
There are 15 separate exponential smoothing
methods.
Each can have an additive or multiplicative error,
giving 30 separate models.
Only 19 models are numerically stable.
Automatic time series forecasting
Exponential smoothing
36
100. Exponential smoothing methods
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
Trend
Component
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
General notation
E T S : ExponenTial Smoothing
Examples:
A,N,N:
A,A,N:
M,A,M:
Simple exponential smoothing with additive errors
Holt’s linear method with additive errors
Multiplicative Holt-Winters’ method with multiplicative errors
Automatic time series forecasting
Exponential smoothing
37
101. Exponential smoothing methods
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
Trend
Component
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
General notation
E T S : ExponenTial Smoothing
Examples:
A,N,N:
A,A,N:
M,A,M:
Simple exponential smoothing with additive errors
Holt’s linear method with additive errors
Multiplicative Holt-Winters’ method with multiplicative errors
Automatic time series forecasting
Exponential smoothing
37
102. Exponential smoothing methods
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
Trend
Component
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
General notation
Examples:
A,N,N:
A,A,N:
M,A,M:
E T S : ExponenTial Smoothing
↑
Trend
Simple exponential smoothing with additive errors
Holt’s linear method with additive errors
Multiplicative Holt-Winters’ method with multiplicative errors
Automatic time series forecasting
Exponential smoothing
37
103. Exponential smoothing methods
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
Trend
Component
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
General notation
Examples:
A,N,N:
A,A,N:
M,A,M:
E T S : ExponenTial Smoothing
↑
Trend Seasonal
Simple exponential smoothing with additive errors
Holt’s linear method with additive errors
Multiplicative Holt-Winters’ method with multiplicative errors
Automatic time series forecasting
Exponential smoothing
37
104. Exponential smoothing methods
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
Trend
Component
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
General notation
Examples:
A,N,N:
A,A,N:
M,A,M:
E T S : ExponenTial Smoothing
↑
Error Trend Seasonal
Simple exponential smoothing with additive errors
Holt’s linear method with additive errors
Multiplicative Holt-Winters’ method with multiplicative errors
Automatic time series forecasting
Exponential smoothing
37
105. Exponential smoothing methods
Seasonal Component
N
A
M
(None)
(Additive) (Multiplicative)
Trend
Component
N
(None)
N,N
N,A
N,M
A
(Additive)
A,N
A,A
A,M
Ad
(Additive damped)
Ad ,N
M
(Multiplicative)
M,N
Ad ,A
M,A
Ad ,M
M,M
Md
(Multiplicative damped)
Md ,N
Md ,A
Md ,M
General notation
Examples:
A,N,N:
A,A,N:
M,A,M:
E T S : ExponenTial Smoothing
↑
Error Trend Seasonal
Simple exponential smoothing with additive errors
Holt’s linear method with additive errors
Multiplicative Holt-Winters’ method with multiplicative errors
Automatic time series forecasting
Exponential smoothing
37
106. Exponential smoothing methods
Innovations state space modelsComponent
Seasonal
Trend
N
A
M
¯ AllComponent
ETS models can (None) (Additive) (Multiplicative)
be written in innovations
N
state space form (IJF, 2002). N,A
(None)
N,N
A
(Additive)
N,M
A,N
A,A
A,M
d
d
d
d
d
d
¯ Additive and multiplicative versions give,M
the
A
(Additive damped)
A ,N
A ,A
A
d
M
Md
same point forecasts but different prediction
(Multiplicative)
M,N
M,A
M,M
intervals. damped) M ,N
(Multiplicative
M ,A
M ,M
General notation
Examples:
A,N,N:
A,A,N:
M,A,M:
E T S : ExponenTial Smoothing
↑
Error Trend Seasonal
Simple exponential smoothing with additive errors
Holt’s linear method with additive errors
Multiplicative Holt-Winters’ method with multiplicative errors
Automatic time series forecasting
Exponential smoothing
37
107. ETS state space models
xt −1
yt
εt
xt
yt+1
εt+1
xt+1
yt+2
εt+2
xt +2
y t +3
εt+3
xt +3
State space model
xt = ( t , bt , st , st−1 , . . . , st−m+1 )
y t +4
Estimation
Optimize L wrt θ = (α, β, γ, φ)
and initial states
x0 = ( 0 , b0 , s0 , s−1 , . . . , s−m+1 ).
Automatic time series forecasting
ε t +4
Exponential smoothing
38
108. ets algorithm in R
Based on Hyndman Khandakar
(IJF 2008):
Apply each of 19 models that are
appropriate to the data. Optimize
parameters and initial values
using MLE.
Select best method using AICc.
Produce forecasts using best
method.
Obtain prediction intervals using
underlying state space model.
Automatic time series forecasting
Exponential smoothing
39
109. ets algorithm in R
Based on Hyndman Khandakar
(IJF 2008):
Apply each of 19 models that are
appropriate to the data. Optimize
parameters and initial values
using MLE.
Select best method using AICc.
Produce forecasts using best
method.
Obtain prediction intervals using
underlying state space model.
Automatic time series forecasting
Exponential smoothing
39
110. ets algorithm in R
Based on Hyndman Khandakar
(IJF 2008):
Apply each of 19 models that are
appropriate to the data. Optimize
parameters and initial values
using MLE.
Select best method using AICc.
Produce forecasts using best
method.
Obtain prediction intervals using
underlying state space model.
Automatic time series forecasting
Exponential smoothing
39
111. ets algorithm in R
Based on Hyndman Khandakar
(IJF 2008):
Apply each of 19 models that are
appropriate to the data. Optimize
parameters and initial values
using MLE.
Select best method using AICc.
Produce forecasts using best
method.
Obtain prediction intervals using
underlying state space model.
Automatic time series forecasting
Exponential smoothing
39
112. Exponential smoothing
400
300
millions of sheep
500
600
Forecasts from ETS(M,A,N)
1960
1970
1980
1990
2000
2010
Year
Automatic time series forecasting
Exponential smoothing
40
113. Exponential smoothing
400
fit - ets(livestock)
fcast - forecast(fit)
plot(fcast)
300
millions of sheep
500
600
Forecasts from ETS(M,A,N)
1960
1970
1980
1990
2000
2010
Year
Automatic time series forecasting
Exponential smoothing
41
115. Exponential smoothing
1.2
0.6
0.8
1.0
fit - ets(h02)
fcast - forecast(fit)
plot(fcast)
0.4
Total scripts (millions)
1.4
1.6
Forecasts from ETS(M,Md,M)
1995
2000
2005
2010
Year
Automatic time series forecasting
Exponential smoothing
43
116. M3 comparisons
Method
MAPE sMAPE MASE
Theta
ForecastPro
ForecastX
Automatic ANN
B-J automatic
17.42
18.00
17.35
17.18
19.13
12.76
13.06
13.09
13.98
13.72
1.39
1.47
1.42
1.53
1.54
ETS
18.06
13.38
1.52
Automatic time series forecasting
Exponential smoothing
44
117. Outline
1 Motivation
2 Forecasting competitions
3 Forecasting the PBS
4 Time series forecasting
5 Evaluating forecast accuracy
6 Exponential smoothing
7 ARIMA modelling
8 Automatic nonlinear forecasting?
9 Time series with complex seasonality
10 Forecasts about automatic forecasting
Automatic time series forecasting
ARIMA modelling
45
118. Time series forecasting
yt−1
yt−2
yt
yt−3
ARIMA model
Autoregression moving
average (ARMA) model
applied to differences.
εt
εt−1
εt−2
Estimation
Compute likelihood L from
ε1 , ε2 , . . . , εT .
Use optimization
algorithm to maximize L.
Automatic time series forecasting
ARIMA modelling
46
122. Auto ARIMA
450
400
350
300
250
millions of sheep
500
550
Forecasts from ARIMA(0,1,0) with drift
1960
1970
1980
1990
2000
2010
Year
Automatic time series forecasting
ARIMA modelling
48
123. Auto ARIMA
Forecasts from ARIMA(0,1,0) with drift
450
400
350
300
250
millions of sheep
500
550
fit - auto.arima(livestock)
fcast - forecast(fit)
plot(fcast)
1960
1970
1980
1990
2000
2010
Year
Automatic time series forecasting
ARIMA modelling
49
124. Auto ARIMA
1.0
0.8
0.6
0.4
Total scripts (millions)
1.2
1.4
Forecasts from ARIMA(3,1,3)(0,1,1)[12]
1995
2000
2005
2010
Year
Automatic time series forecasting
ARIMA modelling
50
125. Auto ARIMA
0.6
0.8
1.0
fit - auto.arima(h02)
fcast - forecast(fit)
plot(fcast)
0.4
Total scripts (millions)
1.2
1.4
Forecasts from ARIMA(3,1,3)(0,1,1)[12]
1995
2000
2005
2010
Year
Automatic time series forecasting
ARIMA modelling
51
126. How does auto.arima() work?
A non-seasonal ARIMA process
φ(B)(1 − B)d yt = c + θ(B)εt
Need to select appropriate orders p, q, d, and
whether to include c.
Algorithm choices driven by forecast accuracy.
Automatic time series forecasting
ARIMA modelling
52
127. How does auto.arima() work?
A non-seasonal ARIMA process
φ(B)(1 − B)d yt = c + θ(B)εt
Need to select appropriate orders p, q, d, and
whether to include c.
Hyndman Khandakar (JSS, 2008) algorithm:
Select no. differences d via KPSS unit root test.
Select p, q, c by minimising AICc.
Use stepwise search to traverse model space,
starting with a simple model and considering
nearby variants.
Algorithm choices driven by forecast accuracy.
Automatic time series forecasting
ARIMA modelling
52
128. How does auto.arima() work?
A non-seasonal ARIMA process
φ(B)(1 − B)d yt = c + θ(B)εt
Need to select appropriate orders p, q, d, and
whether to include c.
Hyndman Khandakar (JSS, 2008) algorithm:
Select no. differences d via KPSS unit root test.
Select p, q, c by minimising AICc.
Use stepwise search to traverse model space,
starting with a simple model and considering
nearby variants.
Algorithm choices driven by forecast accuracy.
Automatic time series forecasting
ARIMA modelling
52
129. How does auto.arima() work?
A seasonal ARIMA process
Φ(Bm )φ(B)(1 − B)d (1 − Bm )D yt = c + Θ(Bm )θ(B)εt
Need to select appropriate orders p, q, d, P, Q, D, and
whether to include c.
Hyndman Khandakar (JSS, 2008) algorithm:
Select no. differences d via KPSS unit root test.
Select D using OCSB unit root test.
Select p, q, P, Q, c by minimising AIC.
Use stepwise search to traverse model space,
starting with a simple model and considering
nearby variants.
Automatic time series forecasting
ARIMA modelling
53
132. M3 conclusions
MYTHS
Simple methods do better.
Exponential smoothing is better than ARIMA.
FACTS
The best methods are hybrid approaches.
ETS-ARIMA (the simple average of ETS-additive
and AutoARIMA) is the only fully documented
method that is comparable to the M3
competition winners.
Automatic time series forecasting
ARIMA modelling
55
133. M3 conclusions
MYTHS
Simple methods do better.
Exponential smoothing is better than ARIMA.
FACTS
The best methods are hybrid approaches.
ETS-ARIMA (the simple average of ETS-additive
and AutoARIMA) is the only fully documented
method that is comparable to the M3
competition winners.
Automatic time series forecasting
ARIMA modelling
55
134. M3 conclusions
MYTHS
Simple methods do better.
Exponential smoothing is better than ARIMA.
FACTS
The best methods are hybrid approaches.
ETS-ARIMA (the simple average of ETS-additive
and AutoARIMA) is the only fully documented
method that is comparable to the M3
competition winners.
Automatic time series forecasting
ARIMA modelling
55
135. M3 conclusions
MYTHS
Simple methods do better.
Exponential smoothing is better than ARIMA.
FACTS
The best methods are hybrid approaches.
ETS-ARIMA (the simple average of ETS-additive
and AutoARIMA) is the only fully documented
method that is comparable to the M3
competition winners.
Automatic time series forecasting
ARIMA modelling
55
136. M3 conclusions
MYTHS
Simple methods do better.
Exponential smoothing is better than ARIMA.
FACTS
The best methods are hybrid approaches.
ETS-ARIMA (the simple average of ETS-additive
and AutoARIMA) is the only fully documented
method that is comparable to the M3
competition winners.
Automatic time series forecasting
ARIMA modelling
55
137. Outline
1 Motivation
2 Forecasting competitions
3 Forecasting the PBS
4 Time series forecasting
5 Evaluating forecast accuracy
6 Exponential smoothing
7 ARIMA modelling
8 Automatic nonlinear forecasting?
9 Time series with complex seasonality
10 Forecasts about automatic forecasting
Automatic time series forecasting
Automatic nonlinear forecasting?
56
138. Automatic nonlinear forecasting
Automatic ANN in M3 competition did poorly.
Linear methods did best in the NN3
competition!
Very few machine learning methods get
published in the IJF because authors cannot
demonstrate their methods give better
forecasts than linear benchmark methods,
even on supposedly nonlinear data.
Some good recent work by Kourentzes and
Crone (Neurocomputing, 2010) on automated
ANN for time series.
Watch this space!
Automatic time series forecasting
Automatic nonlinear forecasting?
57
139. Automatic nonlinear forecasting
Automatic ANN in M3 competition did poorly.
Linear methods did best in the NN3
competition!
Very few machine learning methods get
published in the IJF because authors cannot
demonstrate their methods give better
forecasts than linear benchmark methods,
even on supposedly nonlinear data.
Some good recent work by Kourentzes and
Crone (Neurocomputing, 2010) on automated
ANN for time series.
Watch this space!
Automatic time series forecasting
Automatic nonlinear forecasting?
57
140. Automatic nonlinear forecasting
Automatic ANN in M3 competition did poorly.
Linear methods did best in the NN3
competition!
Very few machine learning methods get
published in the IJF because authors cannot
demonstrate their methods give better
forecasts than linear benchmark methods,
even on supposedly nonlinear data.
Some good recent work by Kourentzes and
Crone (Neurocomputing, 2010) on automated
ANN for time series.
Watch this space!
Automatic time series forecasting
Automatic nonlinear forecasting?
57
141. Automatic nonlinear forecasting
Automatic ANN in M3 competition did poorly.
Linear methods did best in the NN3
competition!
Very few machine learning methods get
published in the IJF because authors cannot
demonstrate their methods give better
forecasts than linear benchmark methods,
even on supposedly nonlinear data.
Some good recent work by Kourentzes and
Crone (Neurocomputing, 2010) on automated
ANN for time series.
Watch this space!
Automatic time series forecasting
Automatic nonlinear forecasting?
57
142. Automatic nonlinear forecasting
Automatic ANN in M3 competition did poorly.
Linear methods did best in the NN3
competition!
Very few machine learning methods get
published in the IJF because authors cannot
demonstrate their methods give better
forecasts than linear benchmark methods,
even on supposedly nonlinear data.
Some good recent work by Kourentzes and
Crone (Neurocomputing, 2010) on automated
ANN for time series.
Watch this space!
Automatic time series forecasting
Automatic nonlinear forecasting?
57
143. Outline
1 Motivation
2 Forecasting competitions
3 Forecasting the PBS
4 Time series forecasting
5 Evaluating forecast accuracy
6 Exponential smoothing
7 ARIMA modelling
8 Automatic nonlinear forecasting?
9 Time series with complex seasonality
10 Forecasts about automatic forecasting
Automatic time series forecasting
Time series with complex seasonality
58
144. Examples
6500 7000 7500 8000 8500 9000 9500
Thousands of barrels per day
US finished motor gasoline products
1992
1994
1996
1998
2000
2002
2004
Weeks
Automatic time series forecasting
Time series with complex seasonality
59
145. Examples
300
200
100
Number of call arrivals
400
Number of calls to large American bank (7am−9pm)
3 March
17 March
31 March
14 April
28 April
12 May
5 minute intervals
Automatic time series forecasting
Time series with complex seasonality
59
147. 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 AM De Livera,
RJ Hyndman, and RD Snyder (2011). “Forecasting
time series with complex seasonal patterns using
exponential smoothing”. Journal of the American
Statistical Association 106(496), 1513–1527.
Automatic time series forecasting
Time series with complex seasonality
60
148. Examples
9000
8000
fit - tbats(gasoline)
fcast - forecast(fit)
plot(fcast)
7000
Thousands of barrels per day
10000
Forecasts from TBATS(0.999, {2,2}, 1, {52.1785714285714,8})
1995
2000
2005
Weeks
Automatic time series forecasting
Time series with complex seasonality
61
149. Examples
500
Forecasts from TBATS(1, {3,1}, 0.987, {169,5, 845,3})
300
200
100
0
Number of call arrivals
400
fit - tbats(callcentre)
fcast - forecast(fit)
plot(fcast)
3 March
17 March
31 March
14 April
28 April
12 May
26 May
9 June
5 minute intervals
Automatic time series forecasting
Time series with complex seasonality
62
150. Examples
Forecasts from TBATS(0, {5,3}, 0.997, {7,3, 354.37,12, 365.25,4})
20
15
10
Electricity demand (GW)
25
fit - tbats(turk)
fcast - forecast(fit)
plot(fcast)
2000
2002
2004
2006
2008
2010
Days
Automatic time series forecasting
Time series with complex seasonality
63
151. Outline
1 Motivation
2 Forecasting competitions
3 Forecasting the PBS
4 Time series forecasting
5 Evaluating forecast accuracy
6 Exponential smoothing
7 ARIMA modelling
8 Automatic nonlinear forecasting?
9 Time series with complex seasonality
10 Forecasts about automatic forecasting
Automatic time series forecasting
Forecasts about automatic forecasting
64
152. Forecasts about forecasting
1
2
3
4
Automatic algorithms will become more
general — handling a wide variety of time
series.
Model selection methods will take account
of multi-step forecast accuracy as well as
one-step forecast accuracy.
Automatic forecasting algorithms for
multivariate time series will be developed.
Automatic forecasting algorithms that
include covariate information will be
developed.
Automatic time series forecasting
Forecasts about automatic forecasting
65
153. Forecasts about forecasting
1
2
3
4
Automatic algorithms will become more
general — handling a wide variety of time
series.
Model selection methods will take account
of multi-step forecast accuracy as well as
one-step forecast accuracy.
Automatic forecasting algorithms for
multivariate time series will be developed.
Automatic forecasting algorithms that
include covariate information will be
developed.
Automatic time series forecasting
Forecasts about automatic forecasting
65
154. Forecasts about forecasting
1
2
3
4
Automatic algorithms will become more
general — handling a wide variety of time
series.
Model selection methods will take account
of multi-step forecast accuracy as well as
one-step forecast accuracy.
Automatic forecasting algorithms for
multivariate time series will be developed.
Automatic forecasting algorithms that
include covariate information will be
developed.
Automatic time series forecasting
Forecasts about automatic forecasting
65
155. Forecasts about forecasting
1
2
3
4
Automatic algorithms will become more
general — handling a wide variety of time
series.
Model selection methods will take account
of multi-step forecast accuracy as well as
one-step forecast accuracy.
Automatic forecasting algorithms for
multivariate time series will be developed.
Automatic forecasting algorithms that
include covariate information will be
developed.
Automatic time series forecasting
Forecasts about automatic forecasting
65
156. For further information
robjhyndman.com
Slides and references for this talk.
Links to all papers and books.
Links to R packages.
A blog about forecasting research.
Automatic time series forecasting
Forecasts about automatic forecasting
66