00 mi piace00 non mi piace

2 visualizzazioni7 pagineMay 28, 2020

© © All Rights Reserved

2 visualizzazioni

00 mi piace00 non mi piace

Sei sulla pagina 1di 7

journal homepage: www.elsevier.com/locate/eswa

Tae Woo Joo 1, Seoung Bum Kim ⇑

School of Industrial Management Engineering, Korea University, Anam-Dong, Seongbuk-Gu, Seoul 136-713, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: Forecasting time series data is one of the most important issues involved in numerous applications in real

Available online 20 January 2015 life. Time series data have been analyzed in either the time or frequency domains. The objective of this

study is to propose a forecasting method based on wavelet ﬁltering. The proposed method decomposes

Keywords: the original time series into the trend and variation parts and constructs a separate model for each part.

ARIMA Simulation and real case studies were conducted to examine the properties of the proposed method

Forecasting under various scenarios and compare its performance with time series forecasting models without wave-

Time series

let ﬁltering. The results from both simulated and real data showed that the proposed method based on

Wavelet transforms

wavelet ﬁltering yielded more accurate results than the models without wavelet ﬁltering in terms of

mean absolute percentage error criterion.

Ó 2015 Elsevier Ltd. All rights reserved.

forms that are localized in both the time and frequency domains

A time series is a set of observations collected over time (Lai (Mallat, 1989) have been proposed (Morettin, 1996, 1997;

et al., 2010). The analysis of time series data is one of the most Mousa, Munib, & Moussa, 2005; Percival & Walden, 1999;

important areas in statistics in both theory and application. The Priestley, 1996). More speciﬁcally, because wavelet basis functions

main objective of time series analysis is to obtain inherent struc- exist over a ﬁnite time limit and are typically irregular and asym-

tural characteristics, such as autocorrelations, trends, and seasonal metric, they are better suited for those time series analyses that

variations, and to use this information to formulate an appropriate exhibit sharp discontinuities and local behavior (Graps, 1995).

mathematical model for analysis and prediction (Anderson, 1971). Many prediction approaches based on wavelet transforms have

In general, time series data can be analyzed in either the time or been developed recently (Chen, 2014; Nguyen, Khosravi,

frequency domains. The most widely used methods in analyzing Creighton, & Nahavandi, 2015). Nguyen et al. (2015) presents a

the time domain include time series regression, decomposition combination of wavelet features with fuzzy standard additive

methods, exponential smoothing, and the Box–Jenkins autoregres- model for medical diagnosis. In general, if a time series is non-

sive integrated moving average (Li, Gai, Kang, Wu, & Wang, 2014). stationary, it is difﬁcult to determine a relevant global model. For

Analyses in the frequency domain are usually conducted for example, Hydrological time series forecasting is a difﬁcult task

periodic and cyclical observations. The methodologies of the fre- because of its complicated nonlinear, non-stationary and multi-

quency domain are based on Fourier transforms that allow us to scale characteristics (Di, Yang, & Wang, 2014). To overcome this

determine the number of frequency components and detect the problem, local models based on wavelet transforms have been pro-

dominant cyclic frequency, all of which are embedded in the time posed (Weigend & Mangeas, 1995; Zhang, Coggins, Jabri, Dersch, &

domain. However, Fourier transforms-based methods have limita- Flower, 2001). In addition, it is known that wavelet transforms

tions in that they require an assumption of stationarity and pro- have the potential to increase the accuracy of time series predic-

duce no information associated with time (Croarkin & Tobias, tions (Ramsey, 1999; Schlüter & Deuschle, 2010). Once forecasting

2006; Popoola, 2007; Tabak and Feitosa, 2010). Further, Fourier models using wavelet transforms succeeded in eliminating noise

transforms do not work well in large data, and thus, they can be before preceding to construct a model (Alrumaih & Al-Fawzan,

implemented only on an interval between 0 and 2p (Tran, 2006). 2002), improvement in overall forecasting performance followed

(Chou, 2014; Ferbar, Čreslovnik, Mojškerc, & Rajgelj, 2009).

Other approaches that use wavelet transforms estimate compo-

⇑ Corresponding author. Tel.: +82 2 3290 3397; fax: +82 2 929 5888.

nents in a structural time series model (Arino, 1995; Sang, 2013;

E-mail addresses: mrblond002@gmail.com (T.W. Joo), sbkim1@korea.ac.kr

Wong, Wai-Cheung, Zhongjie, & Lui, 2003; Zhang et al., 2001).

(S.B. Kim).

1

Tel.: +82 2 3290 3769. Arino (1995) proposed a methodology that used wavelet

http://dx.doi.org/10.1016/j.eswa.2015.01.026

0957-4174/Ó 2015 Elsevier Ltd. All rights reserved.

T.W. Joo, S.B. Kim / Expert Systems with Applications 42 (2015) 3868–3874 3869

Z Z

transforms to decompose time series into long-term trends and

/ðtÞdx ¼ 1 wðtÞdx ¼ 0 /ðtÞ; wðtÞ 2 L2 ðRÞ;

seasonal components. Speciﬁcally, let y = {yt : t = 1, . . ., T} be a time

series set. y is decomposed into two sets z1 = {z1t : t = 1, . . ., T} and

where j and k denote, respectively, the scaling parameter and trans-

z2 = {z2t : t = 1, . . ., T}, that is, y ¼ z1 þ z2 , where z1 and z2 , respec-

lation index. L2 ðRÞ is the space of square integrable real function

tively, represent long-term trends and seasonal behavior. Each

deﬁned on the real line R. The data can be decomposed into the fol-

component is used to construct an appropriate model, and total

lowing wavelet series form:

prediction accuracy can be obtained by aggregating the prediction

results from each component model. However, the generalizability X

n X

J X

n

of this method is questionable because Arino’s method considered yðtÞ ¼ cj;k /j;k ðtÞ þ dj;k wj;k ðtÞ; ð2Þ

only limited cases such as long-term trends and seasonal k¼1 j¼1 k¼1

variations.

where cj;k represents the approximation coefﬁcient at scale j and

Moreover, some methods applied linear or nonlinear predictive

location k; J is the decomposition level; dj;k represents the detail

models to wavelet coefﬁcients (approximation and detail coefﬁ-

coefﬁcient at scale j and location k; and n is the size of the time ser-

cients) at each level, and the ﬁnal prediction was obtained by

ies data. Reconstruction can also be done through Mallet’s pyramid

inverting the predicted wavelet coefﬁcients. Renaud, Starck, and

algorithm (Burt & Adelson, 1983). Fig. 1 shows an example of the

Murtagh (2002) introduced the multiscale autoregressive models

overall process of decomposition and reconstruction by wavelet

that use Haar wavelets and scale coefﬁcients during decomposition.

transforms.

A method that combines nonlinear models with wavelet coefﬁ-

In Fig. 1, cj;k represents the coarse approximation (trend and

cients has also been proposed (Chen, Qian, & Meng, 2013; Hadaś-

seasonality) and dj;k represents detailed information (noise or ran-

Dyduch, 2014; Rocha, Paredes, Carvalho, Henriques, & Harris,

dom ﬂuctuation). The difference between the ﬁrst level approxi-

2010; Soltani, 2002; Wang, 2014). Chen, Nicolis, and Vidakovic

mation coefﬁcients c1;k and original series y yields the detail

(2010) classiﬁed the wavelet coefﬁcients into trend, seasonal, and

coefﬁcients of the ﬁrst level d1;k : To obtain c2;k ; c1;k is approximated

high frequency components and used them to construct forecasting

by a set of basis functions. As can be seen from Fig. 1, when the

models based on exponential smoothing, harmonic regression, and

level becomes higher, cj;k represents the overall pattern of the ori-

autoregressive moving average with exogenous input (ARMAX)

ginal series.

model. However, these methods involve a high computational load

because they considered each level of a series. Moreover, consider-

ing each level of series could cause an overﬁtting problem. 3. Proposed forecasting method

In this paper, we propose a forecasting method that uses wave-

let ﬁltering. Through wavelet transforms, the series is partitioned 3.1. Data decomposition through wavelet ﬁltering

into two parts (trends and variations). We then construct a sepa-

rate model for each part. By this partitioning, the proposed method Our proposed method consists of three steps: (1) wavelet trans-

is especially useful for the time series with a large amount of noise. forms to decompose the data in the trend part (TP) and variation

The potential overﬁtting problem of this approach caused by using part (VP), (2) determination of the optimal decomposition level,

two separate models is addressed by adjusting the decomposition and (3) construction of forecasting models. First, wavelet trans-

levels. forms are conducted to the original time series data to obtain the

The rest of this paper is organized as follows. In Section 2, we same number of coefﬁcients as the size of the data. Further, deno-

brieﬂy introduce the basic concept of wavelet transforms. In ising that shrinks the empirical detail wavelet coefﬁcients toward

Section 3, we present our proposed forecasting approach based zero is performed to remove noise. We used soft thresholding in

on wavelet ﬁltering. Section 4 presents a simulation study to this study (Donoho & Johnstone, 1994). A wavelet ﬁltering method

examine the properties of the proposed method and evaluate its is used to partition the data into the TP and VP (Chang and Yadama,

performance under various scenarios. Section 4.2 presents the 2010). More precisely, the TP can be obtained by setting all the

experimental results from real-life problems. Finally, Section 5 detail coefﬁcients at all levels dj;k ðj ¼ 1; . . . ; JÞ to zero while main-

contains our concluding remarks. taining all coarse approximation coefﬁcients, cj;k ðj ¼ 1; . . . ; JÞ. Simi-

larly, the VP can be obtained by setting all the coarse

2. Wavelet transforms approximation coefﬁcients at the last level, cJ;k to zero while main-

taining all detail coefﬁcients. The TP and VP can be considered the

Wavelets have the advantages of the locality of the analysis and overall trend and variation information in a time series. Fig. 2

their ability to handle multiscale information efﬁciently. Numer- shows how a series can be decomposed into the TP and VP by

ous studies of wavelets have been conducted in the ﬁelds of sig- wavelet ﬁltering. Note that in this paper, we used Daubechies

nal/image processing (Meerwald & Uhl, 2001; Prasad & Iyengar, wavelets as a basis function whose scale and detail wavelets are

1997; Rao & Bopardikar, 1998; Subasi, 2007; Avci & Derya, 2008), shown in Fig. 3.

statistics (Abramovich, Bailey, & Sapatinas, 2000; Antoniadis,

1999; Vidakovic, 1999), and manufacturing processes (Guo, 3.2. Choosing the optimal decomposition level

Linyan, Gang, & Song, 2008; Jeong, Lu, & Wang, 2006; Jin & Shi,

2001; Lada, Lu, & Wilson, 2002; Saravanan & Ramachandran, For modeling, the dataset is divided into training (80%), valida-

2009). Although Fourier transforms give only the frequency infor- tion (10%), and testing (10%) sets. The training set is used to create

mation of given data, wavelet transforms can simultaneously deli- a model, and the testing set is used to evaluate the model. The val-

ver both the time and frequency localizations (Mallat, 1989). idation set is used to select the optimal parameters for the decom-

Wavelet transforms are conducted from wavelet basis functions position level of wavelet transforms. To determine an adequate

that consist of a scale wavelet part /ðtÞ and a detail part wðtÞ: decomposition level, we varied the number of levels for both the

TP and VP from one to six until the model delivered the best accu-

/j;k ðtÞ ¼ 2j=2 /ð2j t kÞ; racy in terms of the minimum mean absolute percentage error

(MAPE). We believe a potential overﬁtting problem by using two

separate models (TP and VP) can be addressed by adjusting the

wj;k ðtÞ ¼ 2j=2 wð2j t kÞ; ð1Þ

decomposition levels in the wavelet transforms.

3870 T.W. Joo, S.B. Kim / Expert Systems with Applications 42 (2015) 3868–3874

model that is deﬁned as follows:

Once the time series is decomposed into the TP and VP at its

optimal decomposition levels, we constructed the forecasting

rd yt ¼ /1 y1 þ /2 y2 þ þ /p ytp þ at h1 at1 h2 at2

model for each part. The forecasting model that we used in this hq atq ð3Þ

T.W. Joo, S.B. Kim / Expert Systems with Applications 42 (2015) 3868–3874 3871

ARIMA (p, d, q), where p, d, and q are, respectively, the order of

the autoregressive, integrated, and moving average parts of the

model. Because ARIMA models require an assumption of stationa-

rity, we used an augmented Dickey–Fuller test to check whether

or not the series is stationary (Dickey & Fuller, 1979). If the test indi-

cates that the series is nonstationary, we ﬁrst difference the series.

The augmented Dickey–Fuller test is implemented for the differ-

enced series; if the series is still nonstationary, we again ﬁrst differ-

ence the differenced series. The above process is repeated until the

series is stationary. For data with seasonality, we used seasonal

ARIMA (SARIMA) models that can take seasonal variations into

account. ARIMA (or SARIMA) models for the TP and VP are sepa-

rately determined based on penalty function statistics such as

Akaike’s Information Criterion (Wang, Kang, Shen, Chang, &

Chung, 2010). Finally, total prediction accuracy can be obtained

by aggregating the prediction results from each of the TP and VP

models.

4. Simulation

proposed method based on wavelet ﬁltering and compared it

with the ARIMA and SARIMA models without wavelet ﬁltering.

In our simulation study, we considered the time series with and

without seasonal variations. Fig. 4 shows eight examples of time

series data that exhibit increasingly linear trends without sea-

sonal variations ((1)–(4)) and with seasonal variations ((5)–(8)).

We generated different noise levels from the normal distributions

with mean zeros and four different standard deviations (r = 5, 10,

20 and 30).

In the examples of seasonality, we generated the dataset by

using a function that combines sine and cosine. The total data size

was 132, with a period of 12. To measure the performance of our

methods, we ran 200 replications and calculated MAPE, which

can be computed by the following equation:

n

1X ^

yi yi 100; Fig. 4. Time series data with linearly increasing trend with different amounts of

MAPE ¼ ð4Þ

n i¼1 yi noise: (1), (5): r = 5; (2), (6): r = 10; (3), (7): r = 20; (4), (8): r = 30.

Table 1 shows the average values of MAPE from 200 experi-

Table 1

ments between ARIMA (or SARIMA) models without wavelet ﬁlter- Comparative results between ARIMA (or SARIMA) and the proposed method for

ing and from the proposed procedure that constructs ARIMA (or simulated datasets. The values inside parentheses indicate the standard deviations of

SARIMA) models with wavelet ﬁltering. The values inside the MAPE.

parenthesis indicate standard deviations. The results showed that Scenario ARIMA (or SARIMA) models ARIMA (or SARIMA) models

for all scenarios but one (scenario 1) the proposed forecasting pro- without wavelet ﬁltering with wavelet ﬁltering

cedure yielded smaller MAPEs for the testing set than the ARIMA (proposed method)

(or SARIMA) models did. We can see that for the datasets with a (1) 2.11 (1.18) 2.12 (0.66)

small amount of noise (scenarios 2, 5, 6), the proposed procedure (2) 4.75 (2.08) 4.29 (1.35)

was more accurate, but not to a signiﬁcant degree. However, for (3) 8.73 (2.76) 7.51 (1.79)

(4) 12.33 (3.40) 11.28 (2.5)

scenarios with a large amount of noise (scenarios 3, 4, 7, 8), the

(5) 1.95 (0.44) 1.59 (0.38)

proposed procedure yielded signiﬁcantly better prediction accu- (6) 3.90 (0.94) 3.07 (0.68)

racy than pure ARIMA (or SARIMA) models. These simulation (7) 7.84 (1.92) 6.09 (1.34)

results clearly demonstrated the usefulness of the proposed proce- (8) 12.28 (3.83) 9.30 (1.91)

dure for forecasting.

4.2. Case study variations. The last three datasets (Net Proﬁt, Passengers, Max

Power) exhibit seasonal variations. To identify the relevant periods

In this section we demonstrated the applicability of the pro- of the time series with seasonal variations, we used Fourier trans-

posed forecasting method by using eight real life datasets. Table 2 forms. The Viscosity dataset contains the daily viscosity of chemi-

shows a set of time series plots of the dataset. The values under the cal product XR-22. DVD contains a weekly sales market, and Towel

name of the dataset indicate the size of the data. The ﬁrst ﬁve data- contains the weekly sales of absorbent paper towels. These three

sets (Viscosity, DVD, Towel, KOSPI, and House) lack seasonal datasets are available in the book ‘‘Forecasting, Time Series, and

3872 T.W. Joo, S.B. Kim / Expert Systems with Applications 42 (2015) 3868–3874

Table 2

A set of time series plots of eight real datasets. The values under the name of the dataset indicate the size of the data.

1 Viscosity (150) 42

5 House (143) 200

40 190

180

38

170

36

160

34

150

32

140

30

130

28 120

26 110

24 100

0 50 100 150 0 50 100 150

2 DVD (161) 90

6 Net Proﬁt (120) 1100

1000

80

900

70

800

60 700

600

50

500

40

400

30 300

200

20 0 20 40 60 80 100 120

0 20 40 60 80 100 120 140 160 180

3 Towel (120) 20

7 Passengers (132) 2.2

x 10

6

2

15

1.8

10

1.6

1.4

5

1.2

0 1

0.8

-5

0 20 40 60 80 100 120 0.6

0 20 40 60 80 100 120 140

x 10

6

8 Max Power (365) 8000

7500

1.55

7000

1.5

6500

1.45

6000

1.4

5500

1.35

5000

1.3

4500

1.25

4000

1.2 0 50 100 150 200 250 300 350

0 50 100 150 200 250 300

Regression: An Applied Approach (Fourth Edition)’’ (Bowerman, (Lee, 2009). The Passengers data exhibits seasonal variations with

O’Connell, & Koehler, 2005). KOPSI contains a series of stock prices a period of 12 and contains the monthly number of passengers at

from the Korea Exchange (KRX, http://www.krx.co.kr). The House Jeju International Airport from January 2003 to December 2013;

dataset contains the monthly house prices for ﬁrst time buyers this dataset can be obtained from the Korea Airports Corporation

from February 2002 to December 2013 in the UK and were (http://www.airport.co.kr). Max Power contains the daily maxi-

obtained from the Ofﬁce for National Statistics website (http:// mum electricity demand in 2013 in South Korea. This dataset con-

www.ons.gov.uk). The Net Proﬁt dataset exhibits a linearly increas- tains a seasonal pattern with the period of 7 and can be obtained

ing trend with seasonal variations (period 12) that is available in from the Electric Power Statistics Information System (http://

the book ‘‘Statistical Methods for Prediction (Second Edition)’’ www.kpx.or.kr).

Table 3

The comparative results of MAPE between the ARIMA (or SARIMA) models with and without wavelet ﬁltering.

Data ARIMA models without wavelet ﬁltering ARIMA models with wavelet ﬁltering (proposed method)

Parameters MAPE (%) Parameters MAPE (%)

Optimal decomposition level TP VP

Viscosity (0, 0, 2) 4.81 3 (2, 0, 0) (2, 0, 1) 5.09

DVD (2, 1, 1) 5.38 3 (2, 1, 1) (2, 0, 2) 4.02

Towel (0 ,1, 1) 6.64 4 (2, 2, 2) (2, 0, 1) 6.03

KOSPI (2, 1, 2) 2.88 5 (2, 2, 2) (2, 0, 1) 2.47

House (1, 1, 1) 3.19 2 (2, 2, 2) (2, 0, 1) 2.44

Net Proﬁt (0, 2, 2) (0, 2, 2)s = 12 6.55 4 (2, 2, 2) (0, 2, 2) (0, 2, 2)s = 12 4.83

Passengers (2, 2, 2) (0, 2, 2)s = 12 13.21 4 (2, 2, 2) (2, 2, 2) (0, 2, 2)s = 12 11.13

Max Power (2, 2, 2) (0, 2, 2)s = 7 11.02 5 (2, 1, 2) (1, 0, 1) (0, 2, 2)s = 7 7.85

T.W. Joo, S.B. Kim / Expert Systems with Applications 42 (2015) 3868–3874 3873

Table 3 shows the comparative results with the real data that Avci, Engin, & Derya, A. (2008). A novel approach for digital radio signal

classiﬁcation: Wavelet packet energy–multiclass support vector machine

were obtained with ARIMA (or SARIMA) models without and with

(WPE–MSVM). Expert Systems with Applications, 34(3), 2140–2147.

wavelet ﬁltering. Our proposed method uses the models with Bowerman, B. L., O’Connell, R. T., & Koehler, A. B. (2005). ‘‘Forecasting, time series,

wavelet ﬁltering. The results demonstrated that in seven of eight and regression: An applied approach’’. South-Western.

cases the proposed approach with the forecasting model with Burt, P. J., & Adelson, E. H. (1983). The Laplacian pyramid as a compact image code.

IEEE Transactions on Communications, 31, 532–540.

wavelet ﬁltering yielded smaller MAPEs than the ARIMA (or SAR- Chang, S. I., & Yadama, S. (2010). Statistical process control for monitoring non-

IMA) models. In particular, in the Passenger and Max Power cases, linear proﬁles using wavelet ﬁltering and B-spline approximation. International

the proposed method signiﬁcantly outperformed the SARIMA Journal of Production Research, 48(4), 1049–1068.

Chen, G. (2014). Automatic EEG seizure detection using dual-tree complex wavelet-

model. Note that various optimal decomposition levels were found. Fourier features. Expert Systems with Applications, 41, 2391–2394.

In summary, these results from real life examples demonstrated Chen, H., Nicolis, O., & Vidakovic, B. (2010). Multiscale forecasting method using

the usefulness and applicability of the proposed method. ARMAX models. Current Development in Theory and Applications of Wavelets, 4,

267–287.

Chen, N., Qian, Z., & Meng, X. (2013). Multistep wind speed forecasting based on

5. Conclusions wavelet and gaussian processes. Mathematical Problems in Engineering, 2013,

1–8.

Chou, Chien-Ming (2014). Application of set pair analysis-based similarity forecast

In this paper we have proposed a forecasting method based on model and wavelet denoising for runoff forecasting. Water, 6, 912–928.

wavelet ﬁltering. Time series data usually comprises many local Croarkin, C., & Tobias, P. (2006). NIST/Sematech e-Handbook of Statistical Methods.

U.S. Commerce Department. Available at <http://www.itl.nist.gov/div898/

bumps and peaks with different scales. Wavelets have the advan- handbook/index.htm>.

tage of the locality of the analysis and their ability to handle mul- Di, C., Yang, X., & Wang, X. (2014). A four-stage hybrid model for hydrological time

tiscale information efﬁciently. The main idea of the proposed series forecasting. PLOS ONE, 9(8), e104663.

Dickey, D. A., & Fuller, W. A. (1979). Distribution of the estimators for autoregressive

method is decomposition of the original data into the TP and VP

time series with a unit root. Journal of the American Statistical Association, 74,

that, respectively, represent the overall trend and variations 427–431.

through wavelet ﬁltering and subsequent construction of separate Donoho, D. L., & Johnstone, I. M. (1994). Ideal denoising in an orthonormal basis

chosen from a library of basis. Comptes Rendus de l’Académie des Sciences Paris,

forecasting models for each of them. A potential overﬁtting

319, 1317–1322.

problem by using two models was addressed by adjusting the Ferbar, L., Čreslovnik, D., Mojškerc, B., & Rajgelj, M. (2009). Demand forecasting

decomposition levels in the wavelet transforms. Technically, to methods in a supply chain: Smoothing and denoising. International Journal of

determine an adequate decomposition level, we changed the num- Production Economics, 118, 49–54.

Graps, A. (1995). An introduction to wavelets. IEEE Computational Science and

ber of levels for both the TP and VP from one to six until we found Engineering, 2(2), 50–61.

the model that minimizes MAPE. In the current study, we selected Guo, X., Linyan, S., Gang, L., & Song, W. (2008). A hybrid wavelet analysis and

the decomposition level somewhat arbitrarily. Nevertheless, both support vector machines in forecasting development of manufacturing. Expert

Systems with Applications, 35(1), 415–422.

the simulation and real life examples demonstrated the usefulness Hadaś-Dyduch, M. (2014). Non-classical algorithm for time series prediction of the

and applicability of the proposed forecasting method. In particular, range of economic phenomena with regard to the interaction of ﬁnancial

the proposed method is useful for situations where seasonality market indicators. Chinese Business Review, 13(4), 221–231.

Jeong, M. K., Lu, J. C., & Wang, N. (2006). Wavelet based SPC procedure for

with a large amount of noise exist. complicated functional data. International Journal of Production Research, 44(4),

There are several interesting directions for future research. A 729–744.

more comprehensive simulation should be conducted to evaluate Jin, J., & Shi, J. (2001). Automatic feature extraction of waveform signals for in-

process diagnostic performance improvement. Journal of Intelligent

the proposed forecasting method in terms of efﬁciency and appli-

Manufacturing, 12(3), 267–268.

cability. Another interesting and important future research issue Lada, E. K., Lu, J. C., & Wilson, J. R. (2002). A wavelet-based procedure for

is the derivation of the variance of the forecasting outcome so as process fault detection. IEEE Transaction on Semiconductor Manufacturing, 15(1),

79–90.

to construct the reliable predication interval. Further, we may

Lai, C. P., Chung, P. C., & Tseng, V. S. (2010). A novel two-level clustering method for

use different wavelet basis functions other than Daubechies wave- time series data analysis. Expert Systems with Applications, 37(9), 6319–6326.

lets used in this study to examine their impact to forecasting accu- Li, G., Gai, Z., Kang, X., Wu, Z., & Wang, Y. (2014). ESPSA: A prediction-based

racy. We believe that the fundamental value of the present study algorithm for streaming time series. Expert Systems with Applications, 41,

6098–6105.

lies in its integration of wavelet transforms and traditional fore- Lee, W. W (2009). Statistical methods for prediction (2nd ed.). Free Academy.

casting methods to increase forecasting accuracy. Mallat, S. G. (1989). A theory for multi resolution signal decomposition: The wavelet

representation. IEEE Transaction on Pattern Analysis and Machine Intelligence, 11,

674–693.

Acknowledgements Meerwald, P., & Uhl, A. (2001). ‘‘A survey of wavelet-domain watermarking

algorithms’’. In Proceedings of SPIE, electronic imaging, security and

watermarking of multimedia contents III, 4314 (pp. 505–516), USA.

The authors sincerely thank the referees for their helpful sug-

Morettin, P. A. (1996). ‘‘From fourier to wavelet analysis of time series’’. In A. Pratt

gestions and comments, which greatly improved the quality of (Ed.), Proceedings of computational statistics (pp. 111–122).

the paper. This research was supported by This research was sup- Morettin, P. A. (1997). Wavelets in statistics. Resenhas, 3, 211–272.

ported by Brain Korea PLUS and Basic Science Research Program Mousa, R., Munib, Q., & Moussa, A. (2005). Breast cancer diagnosis system based on

wavelet analysis and fuzzy-neural. Expert Systems with Applications, 28(4),

through the National Research Foundation of Korea funded by 713–723.

the Ministry of Science, ICT and Future Planning (2013007724). Nguyen, T., Khosravi, A., Creighton, D., & Nahavandi, S. (2015). Classiﬁcation of

healthcare data using genetic fuzzy logic system and wavelets. Expert Systems

with Applications, 42, 2184–2197.

References Percival, D. B., & Walden, A. T. (1999). Wavelet methods for time series analysis.

Cambridge: Cambridge University Press.

Abramovich, F., Bailey, T. C., & Sapatinas, T. (2000). Wavelet analysis and its Popoola, A. O. (2007). Fuzzy-Wavelet Method for Time Series Analysis, Disertasi,

statistical applications. The Statistician, 49, 1–29. Department of Computing, School of Electronics and Physical Sciences,

Alrumaih, R. M., & Al-Fawzan, M. A. (2002). Time series forecasting using wavelet University of Surrey, Surrey.

denoising: An application to Saudi stock index. Journal of King Saud University, Prasad, L., & Iyengar, S. S. (1997). Wavelet analysis with applications to image

Engineering Sciences, 2(14), 221–234. processing. CRC Press.

Anderson, T. W. (1971). The statistical analysis of time series. New York: John Wiley & Priestley, M. B. (1996). Wavelets and time-dependent spectral analysis. Journal of

Sons. Time Series Analysis, 17, 85–104.

Antoniadis, A. (1999). Wavelets in statistics: A review. Journal of the Italian Statistical Ramsey, J. B. (1999). The contribution of wavelets to the analysis of economic and

Society. ﬁnancial data. Philosophical Transactions of the Royal Society of London A, 357,

Arino, M. (1995). ‘‘Time Series Forecasts via Wavelets: An Application to Car Sales in 2593–2606.

the Spanish Market’’. Discussion Paper No. 95–30, Institute of Statistics and Rao, R. M., & Bopardikar, A. S. (1998). Wavelet transforms: Introduction to theory and

Decision Sciences, Duke University. applications. Addison Wesley Longman.

3874 T.W. Joo, S.B. Kim / Expert Systems with Applications 42 (2015) 3868–3874

Renaud, O., Starck, J. L., & Murtagh, F. (2002). ‘‘Wavelet-Based Forecasting of Short Tran, L. (2006). ‘‘From fourier transforms to wavelet analysis: Mathematical

and Long Memory Time Series.’’ Université de Genève, Faculté des sciences concepts and examples’’.

économiques et sociales, Département d’économétrie. Vidakovic, B. (1999). Statistical modeling by wavelets. New York: John Wiley & Sons.

Rocha, T., Paredes, S., Carvalho, P., Henriques, J., & Harris, M. (2010). ‘‘Wavelet-based Wang, C. C., Kang, Y., Shen, P. C., Chang, Y. P., & Chung, Y. L. (2010). Applications of

time series forecast with application to acute hypotensive episodes prediction’’. fault diagnosis in rotating machinery by using time series analysis with neural

In Engineering in medicine and biology society (EMBC), 2010 annual international network. Expert Systems with Applications, 37(2), 1696–1702.

conference of the IEEE, 2403–2406. Wang, J. (2014). A hybrid wavelet transform based short-term wind speed

Sang, Y. F. (2013). Improved wavelet modeling framework for hydrologic time series forecasting approach. The Scientiﬁc World Journal.

forecasting. Water Resources Management, 27(8), 2807–2821. Weigend, A. S., & Mangeas, M. (1995). Nonlinear gated experts for time series:

Saravanan, N., & Ramachandran, K. I. (2009). Fault diagnosis of spur bevel gear box Discovering regimes and avoiding overﬁtting. International Journal of Neural

using discrete wavelet features and decision tree classiﬁcation. Expert Systems Systems, 6, 373–399.

with Applications, 36(5), 9564–9573. Wong, H., Wai-Cheung, I., Zhongjie, X., & Lui, X. (2003). Modelling and forecasting

Schlüter S., & Deuschle, C. (2010). ‘‘Using wavelets for time series forecasting? Does by wavelets, and the application to exchange rates. Journal of Applied Statistics,

it pay off?’’ IWQW, University of Erlangen-Nuremberg. 30, 537–555.

Soltani, S. (2002). On the use of wavelet decomposition for time series prediction. Zhang, B. L., Coggins, R., Jabri, M. A., Dersch, D., & Flower, B. (2001). Multiresolution

Neurocomputing, 48, 267–277. forecasting for futures trading using wavelet decompositions. IEEE Transactions

Subasi, A. (2007). EEG signal classiﬁcation using wavelet feature extraction and a on Neural Networks, 12, 765–775.

mixture of expert model. Expert Systems with Applications, 32(4), 1084–1093.

Tabak, B. M., & Feitosa, M. A. (2010). Forecasting industrial production in Brazil:

Evidence from a wavelet approach. Expert Systems with Applications, 37(9),

6345–6351.

## Molto più che documenti.

Scopri tutto ciò che Scribd ha da offrire, inclusi libri e audiolibri dei maggiori editori.

Annulla in qualsiasi momento.