CS计算机代考程序代写 Bayesian flex Week-4 Exponential Smoothing Methods

Week-4 Exponential Smoothing Methods

Some of the slides are adapted from the lecture notes provided by Prof. Antoine Saure and Prof. Rob Hyndman

Business Forecasting Analytics
ADM 4307 – Fall 2021

Exponential Smoothing Methods

Ahmet Kandakoglu, PhD

04 October, 2021

Outline

• Review of last lecture

• Exponential smoothing

• Simple exponential smoothing

• Models with trend

• Models with seasonality

• Taxonomy of exponential smoothing methods

• Innovations state space models

• Forecasting with exponential smoothing

Fall 2021 ADM 4307 Business Forecasting Analytics 2

Review of Last Lecture

• Transformations and adjustments

• Time series components

• Seasonal adjustment

• Decomposition methods

Fall 2021 ADM 4307 Business Forecasting Analytics 3

Trend and Seasonality Patterns

• Patterns based on Pegel’s (1969) classification

• An inappropriate forecasting model, even when optimized, will be inferior to a

more appropriate model

Fall 2021 ADM 4307 Business Forecasting Analytics 4

The Forecasting Scenario

Fall 2021 ADM 4307 Business Forecasting Analytics 5

The Mean as a Forecasting Tool

• The mean as a forecasting tool:

• Appropriate when the data is stationary

• Unable to capture the data pattern if the time series involves other

components

• There is a need to improve upon the mean as the forecast for the next

period(s)

• Classification of forecasting methods discussed in this lecture:

• Averaging methods (equally weighted observations)

• Exponential smoothing methods (unequal weights)

Fall 2021 ADM 4307 Business Forecasting Analytics 7

Averaging Methods

• The mean:

• Good results if the underlying process has no noticeable trend and no

noticeable seasonality

• Cell A-1 in Pegel’s table

• The larger the number of observations, the more stable it becomes

• Only two values need to be stored as times moves on

Fall 2021 ADM 4307 Business Forecasting Analytics 8

Averaging Methods

• Moving averages:

• MA(𝑘) denotes a moving average forecast of order 𝑘 and 𝑘 MA a moving
average smoother of order 𝑘

• It deals with the latest 𝑘 periods of known data

• It requires more storage because all of the 𝑘 latest observations must be
stored

• It cannot handle trend or seasonality very well, although it can do better

than the mean

Fall 2021 ADM 4307 Business Forecasting Analytics 9

Averaging Methods

• Moving average:

• A forecaster must choose the number of periods in the moving average

• The more observations included in the moving average, the greater the

smoothing effect

• Extreme cases

• 𝑘 = 1 (naïve forecast)

• 𝑘 = 𝑛 (mean forecast)

• Lags behind the pattern by one or more periods

• Exponential smoothing methods are generally superior

Fall 2021 ADM 4307 Business Forecasting Analytics 10

Exponential Smoothing

• An obvious extension to the moving average method

• Exponential smoothing methods are weighted averages of past observations, with

weights decaying exponentially as the observations get older

• The most recent observations usually provide the best guide as to the future

Fall 2021 ADM 4307 Business Forecasting Analytics 11

Exponential Smoothing

• Combine a “level”, “trend” and “seasonal” component to describe a time series.

• The rate of change of the components are controlled by “smoothing parameters”: α, β

and γ respectively.

• Need to choose best values for the smoothing parameters (and initial states).

• α controls the flexibility of the level

• If α = 0, the level never updates (mean)

• If α = 1, the level updates completely (naive)

• β controls the flexibility of the trend

• If β = 0, the trend is linear

• If β = 1, the trend changes suddenly every observation

• γ controls the flexibility of the seasonality

• If γ = 0, the seasonality is fixed (seasonal means)

• If γ = 1, the seasonality updates completely
Fall 2021 ADM 4307 Business Forecasting Analytics 12

ExponenTial Smoothing (ETS) Models

General notation

• Error: Additive (“A”) or multiplicative (“M”)

• Trend: None (“N”), additive (“A”), multiplicative (“M”), or damped (“Ad” or “Md”).

• Seasonality: None (“N”), additive (“A”) or multiplicative (“M”)

Fall 2021 ADM 4307 Business Forecasting Analytics 13

E T S

Error Trend Seasonality

Outline

• Review of last lecture

• Exponential smoothing

• Simple exponential smoothing

• Models with trend

• Models with seasonality

• Taxonomy of exponential smoothing methods

• Innovations state space models

• Forecasting with exponential smoothing

Fall 2021 ADM 4307 Business Forecasting Analytics 14

Simple Exponential Smoothing

• Takes the forecast for the previous period and adjusts it using the forecast error:

𝐹𝑡+1 = 𝐹𝑡 + α 𝑌𝑡 − 𝐹𝑡

• The effect of a large or small α is analogous to that of including a small or a large
number of observations when computing a moving average

• Trails any trend in the actual data, since the most the method can do is adjust the

next forecast for some percentage of the most recent error

• Suitable for forecasting data with no trend or seasonal pattern, although the mean

of the data may be changing slowly over time

Fall 2021 ADM 4307 Business Forecasting Analytics 15

Single Exponential Smoothing

• Weights the most recent observation with a weight α and the most recent
forecast with a weight 1 − α

𝐹𝑡+1 = α𝑌𝑡 + 1 − α 𝐹𝑡

• Reduces any storage problem and requires few computations

Fall 2021 ADM 4307 Business Forecasting Analytics 16

Single Exponential Smoothing

• 𝐹𝑡+1 is a weighted moving average of all the past observations

𝐹𝑡+1 =෍

𝑗=0

𝑡−1

α 1 − α 𝑗 𝑌𝑡−𝑗 + 1 − α
𝑡𝐹1

• The rate at which the weights decrease is controlled by the parameter α (the
sum is approximately equal to one)

Fall 2021 ADM 4307 Business Forecasting Analytics 17

Single Exponential Smoothing

• Simple exponential smoothing has a flat forecast function:

𝐹𝑡+ℎ = 𝐹𝑡+1 = 𝑙𝑡 , ℎ = 2, 3,…

• 𝐹1 plays a role in all forecasts generated by the process. If α is small and/or
the time series is relatively short, its weight will have a noticeable effect on the

resulting forecasts

• Initialization (Since the value of 𝐹1 is not known):

• Use the first observed value (𝑌1) as the first forecast or

• Use the average the first four or five values in the data set or

• Use optimization

Fall 2021 ADM 4307 Business Forecasting Analytics 18

Single Exponential Smoothing

• The unknown parameter α and the initial value 𝑙0 for any exponential
smoothing method can be estimated from the observed data by minimizing the

SSE

• This involves a non-linear minimization problem and the use of an optimization

tool

• An alternative approach is trial and error (and considering other measures of

forecast error)

Fall 2021 ADM 4307 Business Forecasting Analytics 19

ETS(A,N,N): Specifying the Model

A model for SES

ETS(y ~ error(“A”) + trend(“N”) + season(“N”))

• By default, an optimal value for α and 𝑙0 is used.

• α can be chosen manually in trend()

trend(“N”, alpha = 0.5)

trend(“N”, alpha_range = c(0.2, 0.8))

Fall 2021 ADM 4307 Business Forecasting Analytics 20

Example: Algerian Exports

algeria_economy <- global_economy %>% filter(Country == “Algeria”)

fit <- algeria_economy %>% model(ETS(Exports ~ error(“A”) + trend(“N”) + season(“N”)))

report(fit)

Fall 2021 ADM 4307 Business Forecasting Analytics 21

Series: Exports

Model: ETS(A,N,N)

Smoothing parameters:

alpha = 0.84

Initial states:

l[0]

39.5

sigma^2: 35.6

AIC AICc BIC

447 447 453

Example: Algerian Exports

fc <- fit %>% forecast(h = 5)

fc

Fall 2021 ADM 4307 Business Forecasting Analytics 22

# A fable: 5 x 5 [1Y]

# Key: Country, .model [1]

Country .model Year Exports .mean

1 Algeria “ETS(Exports ~ error(\”A\”) + trend(\”N\”) + ~ 2018 N(22, 36) 22.4

2 Algeria “ETS(Exports ~ error(\”A\”) + trend(\”N\”) + ~ 2019 N(22, 61) 22.4

3 Algeria “ETS(Exports ~ error(\”A\”) + trend(\”N\”) + ~ 2020 N(22, 86) 22.4

4 Algeria “ETS(Exports ~ error(\”A\”) + trend(\”N\”) + ~ 2021 N(22, 111) 22.4

5 Algeria “ETS(Exports ~ error(\”A\”) + trend(\”N\”) + ~ 2022 N(22, 136) 22.4

Example: Algerian Exports

fc %>% autoplot(algeria_economy) +

geom_line(aes(y = .fitted), col=“red”, data = augment(fit)) +

labs(y=”% of GDP”, title=”Exports: Algeria”) + guides(colour = “none”)

Fall 2021 ADM 4307 Business Forecasting Analytics 23

Outline

• Review of last lecture

• Exponential smoothing

• Simple exponential smoothing

• Models with trend

• Models with seasonality

• Taxonomy of exponential smoothing methods

• Innovations state space models

• Forecasting with exponential smoothing

Fall 2021 ADM 4307 Business Forecasting Analytics 24

Holt’s Linear Trend Method

• Extends single exponential smoothing to allow forecasting of data with trends

• Forecast

• Level

• Trend

• Choose values for 𝛼, 𝛽, 𝑙0 and 𝑏0 to minimize SSE.

• Non-linear optimization

Fall 2021 ADM 4307 Business Forecasting Analytics 25

ETS(A,A,N): Specifying the Model

ETS(y ~ error(“A”) + trend(“A”) + season(“N”))

• By default, optimal values for 𝛽 and 𝑏0 are used.

• 𝛽 can be chosen manually in trend()

trend(“A”, beta = 0.004)

trend(“A”, beta_range = c(0.0, 0.1))

Fall 2021 ADM 4307 Business Forecasting Analytics 26

Example: Australian Population

aus_economy <- global_economy %>% filter(Code == “AUS”) %>% mutate(Pop = Population/1e6)

fit <- aus_economy %>% model(ETS(Pop ~ error(“A”) + trend(“A”) + season(“N”)))

report(fit)

Fall 2021 ADM 4307 Business Forecasting Analytics 27

Series: Pop

Model: ETS(A,A,N)

Smoothing parameters:

alpha = 1

beta = 0.327

Initial states:

l[0] b[0]

10.1 0.222

sigma^2: 0.0041

AIC AICc BIC

-77.0 -75.8 -66.7

Example: Australian Population

fc <- fit %>% forecast(h = 10)

fc

Fall 2021 ADM 4307 Business Forecasting Analytics 28

# A fable: 10 x 5 [1Y]

# Key: Country, .model [1]

Country .model Year Pop .mean

1 Australia “ETS(Pop ~ error(\”A\”) + trend(\”A\”) ~ 2018 N(25, 0.0041) 25.0

2 Australia “ETS(Pop ~ error(\”A\”) + trend(\”A\”) ~ 2019 N(25, 0.011) 25.3

3 Australia “ETS(Pop ~ error(\”A\”) + trend(\”A\”) ~ 2020 N(26, 0.023) 25.7

4 Australia “ETS(Pop ~ error(\”A\”) + trend(\”A\”) ~ 2021 N(26, 0.039) 26.1

5 Australia “ETS(Pop ~ error(\”A\”) + trend(\”A\”) ~ 2022 N(26, 0.061) 26.4

6 Australia “ETS(Pop ~ error(\”A\”) + trend(\”A\”) ~ 2023 N(27, 0.09) 26.8

7 Australia “ETS(Pop ~ error(\”A\”) + trend(\”A\”) ~ 2024 N(27, 0.13) 27.2

8 Australia “ETS(Pop ~ error(\”A\”) + trend(\”A\”) ~ 2025 N(28, 0.17) 27.6

9 Australia “ETS(Pop ~ error(\”A\”) + trend(\”A\”) ~ 2026 N(28, 0.22) 27.9

10 Australia “ETS(Pop ~ error(\”A\”) + trend(\”A\”) ~ 2027 N(28, 0.29) 28.3

Example: Australian Population

fc %>% autoplot(aus_economy) + labs(y = “Millions”, title= “Population: Australia”)

Fall 2021 ADM 4307 Business Forecasting Analytics 29

Outline

• Review of last lecture

• Exponential smoothing

• Simple exponential smoothing

• Models with trend

• Models with seasonality

• Taxonomy of exponential smoothing methods

• Innovations state space models

• Forecasting with exponential smoothing

Fall 2021 ADM 4307 Business Forecasting Analytics 30

Holt-Winters’ Method

• Extends Holt’s method to capture seasonality

• There are two variations to this method:

• Multiplicative seasonality (preferred when seasonal variations are changing

proportional to the level)

• Additive seasonality (preferred when the seasonal variations are roughly

constant)

Fall 2021 ADM 4307 Business Forecasting Analytics 31

Holt-Winters Additive Method

• Additive seasonality:

• Forecast

• Level

• Trend

• Seasonal

Fall 2021 ADM 4307 Business Forecasting Analytics 32

Holt-Winters Multiplicative Method

• Multiplicative seasonality:

• Forecast

• Level

• Trend

• Seasonal

Fall 2021 ADM 4307 Business Forecasting Analytics 33

Specifying the Model

• Holt-Winters additive method with additive errors: ETS(A,A,A)

ETS(y ~ error(“A”) + trend(“A”) + season(“A”))

• Holt-Winters multiplicative method with multiplicative errors: ETS(M,A,M)

ETS(y ~ error(“M”) + trend(“A”) + season(“M”))

Fall 2021 ADM 4307 Business Forecasting Analytics 34

Example: Domestic Overnight Trips in Australia

# Prepare the data

aus_holidays <- tourism %>% filter(Purpose == “Holiday”) %>% summarise(Trips =

sum(Trips)/1000)

# Fit the model

fit <- aus_holidays %>%

model(

additive = ETS(Trips ~ error(“A”) + trend(“A”) + season(“A”)),

multiplicative = ETS(Trips ~ error(“M”) + trend(“A”) + season(“M”))

)

# Forecast the next 3 years

fc <- fit %>% forecast(h = “3 years”)

Fall 2021 ADM 4307 Business Forecasting Analytics 35

Example: Domestic Overnight Trips in Australia

fc %>% autoplot(aus_holidays, level = NULL) + labs(title=”Australian domestic tourism”,

y=”Overnight trips (millions)”)

Fall 2021 ADM 4307 Business Forecasting Analytics 36

Example: Domestic Overnight Trips in Australia

components(fit)

Fall 2021 ADM 4307 Business Forecasting Analytics 37

# A dable: 168 x 7 [1Q]

# Key: .model [2]

# : Trips = lag(level, 1) + lag(slope, 1) + lag(season, 4) + remainder

.model Quarter Trips level slope season remainder

1 additive 1997 Q1 NA NA NA 1.50 NA

2 additive 1997 Q2 NA NA NA -0.294 NA

3 additive 1997 Q3 NA NA NA -0.670 NA

4 additive 1997 Q4 NA 9.79 0.0211 -0.534 NA

5 additive 1998 Q1 11.8 9.94 0.0425 1.50 0.496

6 additive 1998 Q2 9.28 9.88 0.0245 -0.294 -0.415

7 additive 1998 Q3 8.64 9.75 -0.000841 -0.670 -0.588

8 additive 1998 Q4 9.30 9.77 0.00298 -0.534 0.0884

9 additive 1999 Q1 11.2 9.75 -0.00124 1.50 -0.0976

10 additive 1999 Q2 9.61 9.79 0.00552 -0.294 0.157

# … with 158 more rows

Example: Domestic Overnight Trips in Australia

components(fit) %>% autoplot()

Fall 2021 ADM 4307 Business Forecasting Analytics 38

Outline

• Review of last lecture

• Exponential smoothing

• Simple exponential smoothing

• Models with trend

• Models with seasonality

• Taxonomy of exponential smoothing methods

• Innovations state space models

• Forecasting with exponential smoothing

Fall 2021 ADM 4307 Business Forecasting Analytics 39

Exponential Smoothing Methods

Fall 2021 ADM 4307 Business Forecasting Analytics 40

Taxonomy of Exponential Smoothing

Fall 2021 ADM 4307 Business Forecasting Analytics 41

Outline

• Review of last lecture

• Exponential smoothing

• Simple exponential smoothing

• Models with trend

• Models with seasonality

• Taxonomy of exponential smoothing methods

• Innovations state space models

• Forecasting with exponential smoothing

Fall 2021 ADM 4307 Business Forecasting Analytics 42

State Space Models

• Exponential smoothing methods presented so far generate point forecasts

• There are statistical models that generate the same point forecasts, but can

also generate forecast intervals

• A statistical model is a stochastic (or random) data generating process that

can produce an entire forecast distribution

• Each model consists of a measurement equation that describes the observed

data and some transition equations that describe how the unobserved

components or states (level, trend, seasonal) change over time. Hence these

are referred to as state space models

Fall 2021 ADM 4307 Business Forecasting Analytics 43

State Space Models

• Each model has an observation equation and transition equations, one for each state

(level, trend, seasonal), i.e., state space models.

• Two models for each method: one with additive and one with multiplicative errors, i.e.,

in total 18 models.

• ETS(Error, Trend, Seasonal):

• Error = {A, M}

• Trend = {N, A, Ad}

• Seasonal = {N, A, M}

Fall 2021 ADM 4307 Business Forecasting Analytics 44

State Space Models

Fall 2021 ADM 4307 Business Forecasting Analytics 45

Estimating ETS models

• Smoothing parameters 𝛼, 𝛽, 𝛾 and ∅, and the initial states 𝑙0, 𝑏0, 𝑠0, 𝑠−1, . . , 𝑠−𝑚+1
are estimated by maximizing the “likelihood”.

• The likelihood is the probability of the data arising from the specified model.

• For models with additive errors equivalent to minimizing SSE.

• For models with multiplicative errors, not equivalent to minimizing SSE.

Fall 2021 ADM 4307 Business Forecasting Analytics 46

Model Selection

• Which ETS model is to use?

• The AIC, AICc and BIC can be used here to determine which of the ETS

models is most appropriate for a given time series.

Akaike’s Information Criterion (AIC)

• AIC = −2 log(L) + 2k (where L is the likelihood of the model and k is the total number of

parameters and initial states that have been estimated)

Corrected AIC

• AICc = AIC + (2k(k + 1) / (T − k − 1))

Bayesian Information Criterion (BIC)

• BIC = AIC + k[log(T) − 2]

Fall 2021 ADM 4307 Business Forecasting Analytics 47

The ETS() Function

• Automatically chooses a model by minimizing the AICc

• Can handle any combination of trend, seasonality and damping

• Ensures the parameters are admissible (equivalent to invertible)

• Produces an object of class “ETS”

Fall 2021 ADM 4307 Business Forecasting Analytics 48

Example: Australian Holiday Tourism

aus_holidays <- tourism %>% filter(Purpose == “Holiday”) %>% summarise(Trips = sum(Trips)/1000)

fit <- aus_holidays %>% model(ETS(Trips))

report(fit)

Fall 2021 ADM 4307 Business Forecasting Analytics 49

Series: Trips

Model: ETS(M,N,A)

Smoothing parameters:

alpha = 0.3484054

gamma = 0.0001000018

Initial states:

l[0] s[0] s[-1] s[-2] s[-3]

9.727072 -0.5376106 -0.6884343 -0.2933663 1.519411

sigma^2: 0.0022

AIC AICc BIC

226.2289 227.7845 242.9031

Example: Australian Holiday Tourism

fit %>% augment()

Fall 2021 ADM 4307 Business Forecasting Analytics 50

# A tsibble: 80 x 6 [1Q]

# Key: .model [1]

.model Quarter Trips .fitted .resid .innov

1 ETS(Trips) 1998 Q1 11.8 11.2 0.560 0.0498

2 ETS(Trips) 1998 Q2 9.28 9.63 -0.353 -0.0367

3 ETS(Trips) 1998 Q3 8.64 9.11 -0.468 -0.0514

4 ETS(Trips) 1998 Q4 9.30 9.10 0.201 0.0221

5 ETS(Trips) 1999 Q1 11.2 11.2 -0.0535 -0.00476

6 ETS(Trips) 1999 Q2 9.61 9.39 0.214 0.0227

7 ETS(Trips) 1999 Q3 8.91 9.07 -0.159 -0.0176

8 ETS(Trips) 1999 Q4 9.03 9.17 -0.143 -0.0156

9 ETS(Trips) 2000 Q1 11.1 11.2 -0.105 -0.00940

10 ETS(Trips) 2000 Q2 9.20 9.33 -0.130 -0.0140

# … with 70 more rows

Example: Australian Holiday Tourism

fit %>% forecast(h = 8)

Fall 2021 ADM 4307 Business Forecasting Analytics 51

# A fable: 8 x 4 [1Q]

# Key: .model [1]

.model Quarter Trips .mean

1 ETS(Trips) 2018 Q1 N(13, 0.35) 12.7

2 ETS(Trips) 2018 Q2 N(11, 0.3) 10.9

3 ETS(Trips) 2018 Q3 N(10, 0.31) 10.5

4 ETS(Trips) 2018 Q4 N(11, 0.35) 10.6

5 ETS(Trips) 2019 Q1 N(13, 0.48) 12.7

6 ETS(Trips) 2019 Q2 N(11, 0.43) 10.9

7 ETS(Trips) 2019 Q3 N(10, 0.44) 10.5

8 ETS(Trips) 2019 Q4 N(11, 0.48) 10.6

Example: Australian Holiday Tourism

fit %>% forecast(h = 8) %>% autoplot(aus_holidays)

Fall 2021 ADM 4307 Business Forecasting Analytics 52

Example: Australian Holiday Tourism

components(fit) %>% autoplot()

Fall 2021 ADM 4307 Business Forecasting Analytics 53

Example: Australian Holiday Tourism

gg_tsresiduals(fit)

accuracy(fit)

Fall 2021 ADM 4307 Business Forecasting Analytics 54

# A tibble: 1 x 10

.model .type ME RMSE MAE MPE MAPE MASE RMSSE ACF1

1 ETS(Trips) Training 0.0520 0.428 0.331 0.334 3.45 0.798 0.793 -0.0642

Example: Corticosteroid Drug Sales

h02 <- PBS %>% filter(ATC2 == “H02”) %>% summarise(Cost = sum(Cost))

h02 %>% autoplot(Cost)

Fall 2021 ADM 4307 Business Forecasting Analytics 55

Example: Corticosteroid Drug Sales

h02 %>% model(ETS(Cost ~ error(“A”) + trend(“A”) + season(“A”))) %>% report()

Fall 2021 ADM 4307 Business Forecasting Analytics 56

Series: Cost

Model: ETS(A,A,A)

Smoothing parameters:

alpha = 0.17

beta = 0.00631

gamma = 0.455

Initial states:

l[0] b[0] s[0] s[-1] s[-2] s[-3] s[-4] s[-5] s[-6] s[-7] s[-8] s[-9] s[-10] s[-11]

409706 9097 -99075 -136602 -191496 -174531 -241437 210644 244644 145368 130570 84458 39132 -11674

sigma^2: 3.5e+09

AIC AICc BIC

5585 5589 5642

Example: Corticosteroid Drug Sales

h02 %>% model(ETS(Cost)) %>% report()

Fall 2021 ADM 4307 Business Forecasting Analytics 57

Series: Cost

Model: ETS(M,Ad,M)

Smoothing parameters:

alpha = 0.307

beta = 0.000101

gamma = 0.000101

phi = 0.978

Initial states:

l[0] b[0] s[0] s[-1] s[-2] s[-3] s[-4] s[-5] s[-6] s[-7] s[-8] s[-9] s[-10] s[-11]

417269 8206 0.872 0.826 0.756 0.773 0.687 1.28 1.32 1.18 1.16 1.1 1.05 0.981

sigma^2: 0.0046

AIC AICc BIC

5515 5519 5575

Example: Corticosteroid Drug Sales

h02 %>% model(ETS(Cost)) %>% forecast() %>% autoplot(h02)

Fall 2021 ADM 4307 Business Forecasting Analytics 58

Example: Corticosteroid Drug Sales

fit <- h02 %>%

model(

auto = ETS(Cost),

AAA = ETS(Cost ~ error(“A”) + trend(“A”) + season(“A”))

)

Fall 2021 ADM 4307 Business Forecasting Analytics 59

> fit %>% accuracy()

# A tibble: 4 x 10

.model .type ME RMSE MAE MPE MAPE MASE RMSSE ACF1

1 auto Training 2461. 51102. 38649. -0.0127 4.99 0.638 0.689 -0.0958

2 AAA Training -5780. 56784. 43378. -1.30 6.05 0.716 0.766 0.0258

> fit %>% glance()

# A tibble: 4 x 9

.model sigma2 log_lik AIC AICc BIC MSE AMSE MAE

1 auto 4.59e-3 -2740. 5515. 5519. 5575. 2611438596. 2787191547. 5.01e-2

2 AAA 3.50e+9 -2776. 5585. 5589. 5642. 3224448256. 3441212841. 4.34e+4

Business Forecasting Analytics
ADM 4307 – Fall 2021

Exponential Smoothing Methods

Fall 2021 ADM 4307 Business Forecasting Analytics 60