> QS P{q RYbjbjt+t+ nAAU]0000000$P\\\<"222222jlllb`.`$022222l0022lll20202jT00002jl@lV00j2z[cD>.^Title : Forecasting Demand for Low Earth Orbit Mobile Satellite Service in Korea
Authors : Duk Bin Jun*, Seon K. Kim*, Myoung H. Park**,
Moon S. Bae***, Yoon S. Park*, Young J. Joo***
*KAIST, **Hansung Univ., ***ETRI
Address : Graduate School of Management, KAIST
207-43 Chungryangri-dong, Tongdaemun-gu, Seoul, 130-012, Korea
Telephone : 82-02-958-3634
Fax number: 82-02-958-3604
E-mail : dbjun@cais.kaist.ac.kr, s_bedask@cais.kaist.ac.kr
Forecasting Demand for Low Earth Orbit Mobile Satellite Service in Korea
Abstract
Forecasting a new service diffusion process is critical in designing marketing strategies and analyzing the costs and benefits for service providers. It is very difficult, however, in cases that data are not available.
We suggest the combination of analogy and survey to forecast the demand for Low Earth Orbit(LEO) mobile satellite service in Korea. First, we analyze the diffusion of existing mobile phone service, which is similar to LEO service. The diffusion parameters for mobile phone service are then used in a model for LEO service. A survey was made on two hundred fifty-five subscribers of existing mobile phone service in Korea. We estimate the potential market size of LEO service by applying the logit model to the survey data. Then, we forecast the annual demand for LEO service in Korea from 1998 to 2005. We also derive the price elasticity of market potential of LEO service.
Keywords: analogy, survey, Bass model, logit model
1. Introduction
Forecasting a new service diffusion process is critical for service providers to analyze the costs and benefits. Such a forecast may also help to prevent market failure that may arise after the introduction of a new service and to make long term plans for supply of the service.
Forecasting is very difficult, however, in cases that data are not available. It is impossible to directly use quantitative methods such as the diffusion model of Bass[1]. Therefore, methods such as survey, Analytic Hierarchy Process(AHP), and analogy can be applied to forecast the demand for a new service. In that case, the diffusion parameter estimates can be obtained by using either expert judgment or the diffusion history of analogous products(Mahajan, Muller and Bass[7]). In this paper, we suggest the combination of analogy and survey to forecast the demand for Low Earth Orbit(LEO) mobile satellite service in Korea. We also propose application of the logit model to the survey data in order to estimate the potential market size of LEO service.
In section 2, we discuss the characteristics of LEO service and the similarity between LEO service and mobile phone service. The diffusion process of existing mobile phone service is analyzed in section 3. Section 4 describes a survey that provides information necessary to forecast the demand for LEO service. In section 5, we propose a forecasting method in which the logit model is applied to the survey data. We also forecast the demand for LEO service in Korea. In section 6, the price elasticity of the market potential of LEO service is derived. Section 7 presents discussion and concluding remarks.
2. Characteristics of LEO service and similarity between LEO service and mobile phone service
LEO satellite communication system consists of a large number of satellites in low orbits. It may provide real-time communication services to mobile and stationary users anywhere on the globe; it will offer voice communication, data transfer, positioning, paging, and messaging services. In this paper, we focus on voice communication among these services. Using LEO service, it is possible to communicate with a person wherever he is; without hindrance of geographical features of the earth. A subscriber of LEO service can communicate by his terminal and personal number even when he is in a foreign country. On the contrary, this is not possible for a subscriber of existing mobile phone service. The price of LEO service is much higher than that of mobile phone service. Hence, it is anticipated that a dual mode terminal will be developed; such a terminal will provide both mobile phone service and LEO service. In this scenario, a caller subscribes to both services and chooses one or the other on a call-by-call basis. Although there are some differences in their coverage and characteristics, both are personal mobile communication services that link the areas where PSTN service is insufficient or not offered at all. Bayus[3] suggested a segmentation scheme in which a new product can be associated with an appropriate group of similar products and thus their demand parameters using early estimates of price and cost. The price and cost of the two services considered here are very high in the early stage of their life cycle and will decrease at high speed with the development of technology. From the similarity, we can infer that both will be included in the same segment and their respective demands will increase at the same rate. Hence, we assume that the diffusion parameters for LEO service are the same as those for existing mobile phone service.
3. Diffusion of mobile phone service
The diffusion model of Bass[1] has been widely applied to various products to explain the diffusion process of product sales. When the sales during period t are defined as EMBED Equation.2 , the cumulative sales until time t-1 as EMBED Equation.2 , and market potential as N, the Bass model can be represented as follows:
EMBED Equation.2 . (1)
The parameter, p, represents the innovation effect produced by non-buyers, EMBED Equation.2 . The parameter, q, represents the imitation effect produced by the interaction of the buyers, EMBED Equation.2 , and the non-buyers, EMBED Equation.2 .
Because there are not available data, the usual estimation procedures can not be implemented directly to forecast the demand for LEO service. Therefore, we analyze the diffusion process of existing mobile phone service, which is analogous to LEO service. The Bass model was applied to the annual demand for estimation of p and q in mobile phone service in Korea from 1984 to 1993. Estimates of parameters may be unreasonable if the data are not sufficiently rich. Hence, market size, N, was estimated from a survey that the Research & Research company had made on the non-subscribers of mobile phone service in Korea in 1994. The respondents who presented the intention to use mobile phone service in the future occupied about 40.6% of the sample. From that, the market potential of existing mobile phone service was estimated to be about 12 million in 1994. It was also assumed to increase at the same rate of population growth. Given the market potential, the Bass model is linear in the parameters as shown in eq. (1). The OLS estimation results are summarized in table 1.
4. Survey
In a very competitive society, services that can not satisfy customers will not survive in the market. In such an environment, it is important to assess the potential market based on customers need. A survey is a useful method for understanding the need for telecommunication services and the customers behavior relevant to the use of these services.
We estimate the market size of LEO service by using a survey made on the mobile phone users, who are expected to be the probable adopters of LEO service. The most recent survey in Korea took place in the fall of 1994. Randomly selected 1,000 subscribers of existing mobile phone service were interviewed or mailed. The number of respondents was 255. The sample represents various business categories and geographical regions. We asked questions relevant to the use of mobile telecommunication services and an individuals intention to subscribe to LEO service.
5. Forecast model and results
In this section, we develop a forecast model based on the survey data and forecast the demand for LEO service in Korea.
The utilities that the ith individual will obtain by adopting or not adopting LEO service are denoted as EMBED Equation.2 and EMBED Equation.2 , respectively. The superscripts, (L) and (M) represent LEO and mobile phone service because an individual who dose not subscribe to LEO service remains in existing mobile phone service market. The utilities are divided into two additive parts as follows:
EMBED Equation.2 , (2)
EMBED Equation.2 . (3)
EMBED Equation.2 and EMBED Equation.2 are assumed to be deterministic functions. They are related to the factors considered by the individual in deciding whether to adopt the service. The error terms, EMBED Equation.2 and EMBED Equation.2 are random from the observational perspective. An individual will subscribe to LEO service if the utility obtained by subscribing is greater than that obtained when not subscribing. To simplify the model, we assume that the error terms follow the extreme value distribution(see Ben-Akiva and Lerman[4]). Then, the probability that the ith individual subscribes to LEO service is as follows:
EMBED Equation.2 . (4)
The deterministic terms, EMBED Equation.2 and EMBED Equation.2 are defined as a linear function of various explanatory variables, EMBED Equation.2 and EMBED Equation.2 , which are the observed attributes that affect choice. In this case, the exogenous variables represent the need for using mobile telecommunication service, the need for having a personal number in foreign countries, additional costs a potential subscriber is willing to incur, current monthly spending on mobile phone service during business travel, monthly cost for using LEO service and mobile phone service, income, unsatisfactory experience in using existing mobile phone service, etc. Then, the subscription probability is as follows:
EMBED Equation.2 . (5)
For maximum likelihood estimation of the unknown parameters, the likelihood function, EMBED Equation.2 , is defined as follows:
EMBED Equation.2 , (6)
where the indicator variable, EMBED Equation.2 , is defined as one if the intention to use the service measured on seven-point scale is equal to seven and is defined as zero otherwise.
The specification of this model is summarized in table 2. The two columns correspond to LEO service and mobile phone utilities and the rows correspond to the six coefficients labeled EMBED Equation.2 through EMBED Equation.2 . The entries in the table define the variables that enter into the model. Entries in the last column are the coefficient estimates. At first, many variables relevant to mobile telecommunication service were considered. By removing insignificant variables, statistically significant variables were finally selected. Estimates were obtained by maximizing the likelihood function in eq. (6). Their signs are as expected and we can make several inferences about them. The need for mobile telecommunication and the need for having personal number in foreign countries were each measured on a scale of one to seven. An individual with a greater need for mobile telecommunication and a personal number should be more likely to subscribe to LEO service. Therefore, by multiplying these two variables, we used it as an explanatory variable. We also see that a person is more likely to subscribe to LEO service if he is willing to spend more in addition to current monthly spending to mobile phone service during business travel. The costs for terminal purchase and the fees for subscription, basic service, monthly calls were aggregated into a monthly cost. The definite plan for each price component of LEO service and mobile phone service presented in survey are explained in table 2. Utility decreases if the monthly costs for using either LEO service or mobile phone service increases. Subscription probability increases with personal income. If a potential subscriber has an unsatisfactory experience using mobile phone service, he will be more likely to subscribe to LEO service.
We can infer that the potential non-subscribers will subscribe to LEO service by an average subscription probability, EMBED Equation.2 , which is 0.47. As a result, the market size for LEO service is derived as follows:
EMBED Equation.2 . (7)
EMBED Equation.2 and EMBED Equation.2 are the market potential of LEO service and mobile phone service, respectively. Given the potential market size, we can forecast the annual demand for LEO service using the diffusion parameters for mobile phone service. Figure 1 shows the demand for LEO service.
6. Price elasticity
Price elasticity measures the responsiveness of demand to changes in price. The price for LEO service can change as a result of pricing strategies or of technology development. The price elasticity of the market potential can be derived by using a well-known property of the logit model. The derivation is explained in detail in the appendix.
In this section, we derive, as an example, the price elasticity of the market potential where price represents the fee per minute. The example discussed here is the point elasticity. The subscription probability of the ith individual, EMBED Equation.2 , is a function of the fees for monthly calls included in the monthly cost for using LEO service as shown in eq. (5). The fees for monthly calls are calculated by multiplying the length of calls by the fee per minute. If we assume that the call length is unchanged, the relevant elasticity of the ith individuals subscription probability can be derived as follows:
EMBED Equation.2 , (8)
where EMBED Equation.2 is the relevant coefficient as shown in table 2; EMBED Equation.2 and EMBED Equation.2 represent the call length and the fee per minute, respectively. We can see from eq. (8) that the elasticity is greater for a person using LEO service for a longer period. Because the market potential of LEO service is proportional to the average subscription probability, the elasticity of market potential is calculated as follows:
EMBED Equation.2 . (9)
This is simply an average of the individual levels of elasticity weighted by the subscription probabilities. The fee per minute elasticity of the market potential of LEO service is about -0.097. That is, if the fee per minute increases 1% from present level of 1,600 to 1,616 won per minute, the market potential of LEO service decreases 0.097%.
7. Discussion and conclusions
In this paper, we applied the diffusion model of Bass to existing mobile phone service in order to estimate the diffusion parameters. Then, we analyzed the survey data to estimate the market potential of LEO service. The subscription probability was estimated by fitting the survey data to the logit model. The market potential of LEO service was derived from the subscription probabilities. Given the market potential, we forecasted the demand for LEO service by using the diffusion parameters of existing mobile phone service. The price elasticity of the market potential was also derived. Using the point elasticity is one way to predict changes due to modifications in the independent variables. It is useful when the changes take place in the neighborhood of the point. When there are some changes in variables, we can predict the probability that individual i will choose LEO service by recalculating the utilities. If the price of LEO service increases twofold and that of existing mobile phone service decreases half, then the average subscription probability becomes 0.39 from 0.47 that was the estimated value in the original plan for the price. As the price of LEO service increases, the average probability decreases.
When we made the survey, we had a bright future from the standpoint of national economy. However, the financial situation of Korea has become worse and the real income of people has decreased when compared with then. The price of LEO service was calculated at the exchange rate of about 800 won to the U.S. dollar in 1994. There has been a great raise in the foreign exchange rate since then. In these circumstances, the price of LEO service in Korea will be much higher than we expected. The forecast results of this paper were derived by using the survey made about four years ago and the conditions of those days were very different from those of the present. Therefore, the changes must be considered and great care should be taken when using the results.
References
[1] F. Bass, A New Product Growth Model for Consumer Durables, Management Science 15 (1969) 215-227.
[2] F. Bass, The Adoption of a Marketing Model: Comments and Observations, in: Innovation Diffusion of New Product Acceptance, eds. Vijay Mahajan and Yoram Wind (Cambridge, MA: Ballinger Publishing Company, 1986).
[3] B. Bayus, High-Definition Television : Assessing Demand Forecasts for a Next Generation Consumer Durable, Management Science 39 (1993) 1319-1333.
[4] M. Ben-Akiva and S. Lerman, Discrete Choice Analysis : Theory and Application to Travel Demand, (MIT Press, 1991).
[5] S. K. Kim, A Study on the Method for Forecasting the Demand of a New Telecommunication Service, Master Thesis, KAIST (1995)
[6] V. Mahajan and E. Muller, Innovation Diffusion and New Product Growth Models in Marketing, Journal of Marketing 43 (1979) 55-68.
[7] V. Mahajan, E. Muller and F. Bass, New Product Diffusion Models in Marketing : A Review and Directions for Research, Journal of Marketing 54 (1990) 1-26.
[8] V. Mahajan and R. Peterson, Innovation Diffusion in a Dynamic Potential Adopter Population, Management Science 24 (1978) 1589-1597.
[9] V. Mahajan and S. Sharma, Simple Algebraic Estimation Procedure for Innovation Diffusion Models of New Product Acceptance, Technological Forecasting and Social Change 30 (1986) 331-346.
[10] D. C. Schmittlein and V. Mahajan, Maximum Likelihood Estimation for an Innovation Diffusion Model of New Product Acceptance, Marketing Science 1 (1982) 57-78.
[11] V. Srinivasan and C. H. Mason, Nonlinear Least Squares Estimation of New Product Diffusion Models, Marketing Science 5 (1986) 169-178.
[12] F. Sultan, J. Farley and D. Lehmann, A Meta-Analysis of Applications of Diffusion Models, Journal of Marketing 27 (1990) 70-77.
[13] R. Thomas, Estimating Market Growth for New Products: An Analogical Diffusion Model Approach, Journal of Product Innovation Management 2 (1985) 45-55.
Appendix
The elasticity of the ith individuals subscription probability of the fee per minute can be derived as follows:
EMBED Equation.2
EMBED Equation.2 ,
where EMBED Equation.2 is the relevant coefficient; EMBED Equation.2 and EMBED Equation.2 are the length of call and the fee per minute, respectively.
The elasticity of market potential can be derived as follows:
EMBED Equation.2
EMBED Equation.2 .
Table 1. Results for fitting the Bass model to existing mobile phone service in Korea
ParameterEstimatep0.00041 (0.3609)q0.72296 (0.0001) EMBED Equation.2 GivenThe value in parenthesis denotes the p-value.
EMBED Equation.2 : the market potential of existing mobile phone service at time t
EMBED Equation.2 : the number of population who is 20 years old or over
EMBED Equation.2 : the number of cumulative subscribers of mobile phone service at time t-1
The market potential is assumed to increase at the same rate of population growth.
EMBED Equation.2
Table 2. Results of the application of the logit model to LEO service in Korea
LEO service utility, EMBED Equation.2 Mobile phone service utility, EMBED Equation.2 Estimate EMBED Equation.2 10-3.4573
(0.0002) EMBED Equation.2 interaction of need for mobile telecommunication and need for having personal number in foreign countries00.0265
(0.0402) EMBED Equation.2 costs a potential subscriber is willing to incur in addition to current monthly spending on mobile phone service during business travel03.449E-6
(0.0306) EMBED Equation.2 monthly cost for using LEO servicemonthly cost for using mobile phone service-2.8E-6
(0.0576) EMBED Equation.2 income02.332E-7
(0.0123) EMBED Equation.2 0unsatisfactory experience
in using mobile phone service-0.2383
(0.0515)Testing Global Null Hypothesis : BETA=0
Criterion Intercept Only Intercept and Covariates Chi-Square for Covariates
AIC 255.030 228.909 .
SC 258.240 248.166 .
-2LOG L 253.030 216.909 36.121 with 5 DF (0.0001)
Score . . 31.739 with 5 DF (0.0001)The value in parenthesis denotes the p-value.
In the mobile phone utility,
costs for terminal purchase = 600,000 won, fee for subscription = 732,000 won
fee for basic service = 27,000 won,
the ith individuals fee for monthly calls = EMBED Equation.2 ,
The fee for monthly calls is individually asked in the survey.
the ith individuals monthly cost for using mobile phone service
= EMBED Equation.2
In the LEO service utility,
costs for terminal purchase = 1,200,000 won, fee for basic service = 16,000 won
fee per minute = 1,600 won per minute,
call length of the ith individual = EMBED Equation.2 ,
The call length is estimated from the fee for monthly calls asked in survey.
the ith individuals monthly cost for using LEO service
= EMBED Equation.2
The costs for terminal purchase and the fee for subscription are converted to monthly cost by multiplying 0.01.
Figure 1. Demand for LEO service in Korea
EMBED Excel.Chart.5 \s
PAGE
PAGE 1
*+,-VWjklm./BCDEǼwkj26
UVmHnHj26
CJEHUj26
UVmHnHj26
CJEHUj26
UVmHnHj26
CJEHUj26
UVmHnHj26
CJEHUj26
UVmHnHj26
CJEHUj26
CJUVmHnH
jCJUCJPJCJQJCJ&QR\]x$% d$d$dQR\]x$% z{<=ef+,9: """"y#%%@%c%d%''( (++%+&+++++,344446 6 6!6~799::;;<<l=m===gBhEtEuEEFFGG=HH dz{<=ef+,9: """"y#%%@%c% d4$7$8$d
d########$$$$$$%%1%2%3%4%@%A%T%U%V%W%d%e%x%y%z%{%%%%%Ĺ뭢떋thj H7
UVmHnHj4
CJEHUj4
UVmHnHjPi6
CJEHUjPi6
UVmHnHj"4
CJEHUj"4
UVmHnHj5
CJEHUj5
UVmHnHj5
CJEHUj5
UVmHnH6CJCJ
jCJUj26
CJEHU&c%d%''( (++%+&+++++,344446 6 6!6~799::;d%%%=&>&Q&R&S&T&X&Y&l&m&n&o&''''((((9(:(M(N(O(P(U(V(i(j(k(l(((((Ƚ뭢떋thj4?6
UVmHnHj5
CJEHUj5
UVmHnHj
5
CJEHUj
5
UVmHnHj4
CJEHUj4
UVmHnH6CJj,j6
CJEHUj,j6
UVmHnHjL4
CJEHUjL4
UVmHnHCJ
jCJUj H7
CJEHU&(((((((((++++++++++++++++++++,, ,
,P-Q-d-e-f-g-Ƚ뱦뚏xlajx5
CJEHUjx5
UVmHnHj;6
CJEHUj;6
UVmHnHjᴬ6
CJEHUjᴬ6
UVmHnHj6
CJEHUj6
UVmHnHjY6
CJEHUjY6
UVmHnHjl6
CJEHUjl6
UVmHnHCJ
jCJUj4?6
CJEHU&g-p-q-----94:4M4N4O4P4444444444444444444^8_8n8o8888899shj6
CJEHUj6
UVmHnH6CJjVÇ5
CJEHUjVÇ5
UVmHnHj5
CJEHUj5
UVmHnHj@6
CJEHUj@6
UVmHnHj6
CJEHUj6
UVmHnHj5
CJEHUj5
UVmHnH
jCJUCJ(999
:::
: :!:4:5:6:7:i:j:}:~:::::::::;;<<<<@@fBgBuEEEEGFGHH_I`II{vqvqvqvqvCJPJCJPJ>*CJ6CJj6
CJEHUj6
UVmHnHjd6
CJEHUjd6
UVmHnHj6
CJEHUj6
UVmHnHjd6
CJEHUjd6
UVmHnHj6
CJEHUj6
UVmHnH
jCJUCJ,;;<<l=m===gBhEtEuEEFFGG=HH`IIJJKK[LLMMzM{M d4$7$8$dH`IIJJKK[LLMMzM{MMMM\N]NNNNN)O*O4O=O>O@OQOROTOeOfO~OOOOOOOP]PPQ*Q+Q{Q|Q}Q~QQQQQRRRRRR1RRRRRRRNSPSYSbScS{SSSSSSSSSTTT(T*TDTbTjTsTtTTTT*UdU
^IIJJZL[LLLMMM M{M|MMMMMMMMMMMMMMMMMMMMNNNNNNNNNNNNŸ~sjo6
CJEHUjo6
UVmHnHjy=6
CJEHUjy=6
UVmHnHjo6
CJEHUjo6
UVmHnHjC6
EHUmHnHjC6
UVmHnHjz=6
CJEHUjz=6
UVmHnH
jCJU6CJCJCJPJCJPJ,{MMMM\N]NNNNN)O*O4O=O>O@OQOROTOeOPP$$cWH$$cqH$d$$ddNNNNNNNNNNNNfOgOzO{O|O}OOOOOOOPP#P$P%P&P]P^PqPrPsPtPQQ&Qޟވ}qfj?܃5
CJEHUj?܃5
UVmHnHj;܃5
CJEHUj;܃5
UVmHnHjۃ5
CJEHUjۃ5
UVmHnHj;5
CJEHUj;5
UVmHnHCJPJj?6
CJEHUj?6
UVmHnHCJ
jCJUjw=6
CJEHUjw=6
UVmHnH&eOfO~OOOOOOOP]PPQ*Q+Q|d
&FVd
&Fd
&Fd$$cH$$cY4H$d$$$cZ4H&Q'Q(Q)Q+Q,QQQQQQQQQQQQQQQQRRRRR-R.R/R0RRRRRRRcSdSwSޟވ}qfjF06
CJEHUjF06
UVmHnHjG06
CJEHUjG06
UVmHnHjH06
CJEHUjH06
UVmHnHjI06
CJEHUjI06
UVmHnHjJ06
CJEHUjJ06
UVmHnHCJPJCJ
jCJUj5
CJEHUj5
UVmHnH&+Q{Q|Q}Q~QQQQQRRRRRR1RRRRRRRӸT$$c4x(#$$cQx(#$d$$dRNSPSYSbScS{SSSSSSSSSTTT(T$$c4x(#$$c4x(#$$c4x(#$d$wSxSySzSSSSSSSTT$T%T&T'TVVVVVVVWDWFWWWWWWWCXDXTXUXhXiXjXkXXެޠެމ~ެrgj2"5
CJEHUj2"5
UVmHnHjz5
CJEHUjz5
UVmHnHj5
CJEHUj5
UVmHnH6CJjC06
CJEHUjC06
UVmHnHjD06
CJEHUjD06
UVmHnHCJ
jCJUjE06
CJEHUjE06
UVmHnH((T*TDTbTjTsTtTTTT*UdUUUUUU $$c4(#$$cm4(#d$$$c4x(#$d$dUUUUUU,VIVVVWAWWWW X0XmXXXY
Y}YYYYYYYYYYYYYYY
$U,VIVVVWAWWWW X0XmXXXY
Y}YYYYY$Bd4$7$8$
Bd4$7$8$$dd^d
&Fd
&Fd
&FdXXXXY Y
YYYYYYYYYYYYYYYYYYYYYY0JmHnH0J
j0JUjO_5
UjO_5
UV jUj1"5
CJEHUj1"5
UVmHnH
jCJUCJ6CJYYYYYYYYYYYYY
Bd4$7$8$h&`#$8 0
000&P P. A!"#p$%200
&P P. A!"#p$%
!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGIJKLMNORVqUXWZY\[_]^`abdcefhgjiklmnoprstvuxwzy{|}~Root Entry F?@ABCEHKMNOPQRSUZ\]^_`abcdegjmopqrsuz|}~L4@4 .1&`&MathTypePTimes New Roman- 2
`@S Times New RomanA- 2
t>
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q}JJ}J
StL{{hhCompObj fObjInfo
Equation Native 4_919155369
F-bD`bDOle
PIC
LMETACompObjf{{ .1@@&&MathTypeP0~- 2
Y 0~- 2
tI Symbol- 2
4-{ 0~- 2
1
&
Systems- 0
FMicrosoft Equation 2.0DS EquationEquation.29q zJDJ0{J
Yt-1L
.ObjInfoEquation Native <_919155368FubDZbDOle
PIC
LMETACompObj0fObjInfo21 &@&MathType "-@@<
Symbol-2
(Symbol-2
Y)0- 2
S 2
-p 2
q 2
Y 2
xN 2
iN 2
Y 0~- 2
tI 2
tI 2
tISymbol- 2
1= 2
+ 2
2
2
2
F
2
F
2
F
2
!
- Symbol- 2
3 -{ 2
]-{ 0~- 2
1 2
1
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q׀zJ<J8{J
St
=p+qYt-1
N()N-Yt-1
()LEquation Native 3_919155367'FebDqbDOle
6PIC
7L S .1@&c&MathTypeSymbol-2
,(Symbol-2
)0- 2
N 2
6Y 0~- 2
tISymbolMETA9CompObjDfObjInfoFEquation Native GP- 2
Q- Symbol- 2
-{ 0~- 2
1
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q4zJDJ0{J
N-Yt-1
()L{{hh{{ .1@@&&MathTypeP0~- 2
Y 0~- _919155366"F"bD@5bDOle
IPIC
!$JLMETAL2
tI Symbol- 2
4-{ 0~- 2
1
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q zJDJ0{J
Yt-1CompObj#%TfObjInfo&VEquation Native W<_919155365 )F0bDbDOle
XPIC
(+YLMETA[CompObj*,ffL S .1@&c&MathTypeSymbol-2
,(Symbol-2
)0- 2
N 2
6Y 0~- 2
tISymbol- 2
Q- Symbol- 2
-{ 0~- 2
1
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qObjInfo-hEquation Native iP_897552159Xm0FbD0@bDOle
k4zJDJ0{J
N-Yt-1
()LN|N .1`&
PIC
/2lLMETAnpCompObj13tfObjInfo4v&MathTypeP0- 2
U 0- 2
@i82
T(L)\\
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q$|J`J<}J
Ui(L) Equation Native w@_8975702617FbDbDOle
xPIC
69yLL| .1`@&
&MathTypeP0- 2
U 0- 2
@i82
T(M)\\
&
META{pCompObj8:fObjInfo;Equation Native @ Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q$|J`J<}J
Ui(M) LC|_882022690>F@bDbDOle
PIC
=@LMETAlC 2 .1` &
&MathTypeP0- 2
U 2
\V 0- 2
@i82
T(L)\\ 2
\i82
(L)\\ 2
i82
(L)\\Symbol- 2
b= 2
+Symbol- 2
e
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q\|JhJD}J
Ui(L)
=Vi(L)
+ei(L)CompObj?AfObjInfoBEquation Native x_921397584EFJbD?@ABCDEFGHIKNOPQTVWXY[`bcdefghijklmnoqtuxz{|}~&MathTypeP0- 2
V 0- 2
i82
"(L)\\
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q$|JhJD}J
Vi(L) _897554188vFbD@nzbDOle
PIC
ux LMETApLN|N .1`&
&MathTypeP0- 2
V 0- 2
i82
$(M)\\
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q$|JhJD}J
Vi(M) L|CompObjwyfObjInfozEquation Native @_917192500}FbDbDOle
PIC
|LMETACompObj~f .1`&
&MathTypeP0- 2
6x 0-
2
%ik8x 2
L 0- 2
>(\ 2
,)\
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q(|JhJD}J
xik(L)Lq|q .ObjInfo!Equation Native "D_919582828F,bD0lAbDOle
$PIC
%LMETA'CompObj.fObjInfo01` &
&MathTypeP0- 2
6x 0-
2
#ik8x 2
M 0- 2
<(\ 2
)\
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q(|JhJD}J
xik(M)L*
* | .1 &%Equation Native 1D_916674990FХ4bDP&MathType0- 2
L 2
P 2
( 2
P P 2
) 0~- 2
i8 2
I@ 2
=
i8 2
;I@ 2
5ri8`0~- 2
7Ni( 2
7i(Symbol- 2
h= 2
/- Symbol- 2
-{Symbol- 2
0~- 2
71 0~- 2
^1
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q|zJDJ0{J
L=PiIi
(1-Pi
)1-Iii
L`{hObjInforEquation Native s_917444667F$^bD0bDOle
vPIC
wLMETAyXCompObjfObjInfo`{ .1@@&&MathTypeP0~- 2
Io 0~- 2
i8
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qzJDJ0{J
IiL .1`&G@Equation Native 4_897554552F:bDxbDOle
PIC
LMETACompObjfObjInfoEquation Native (&MathType0Symbol- 2
!a
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q|JhJD}J
a_897554571.FPBbDAcDOle
PIC
LMETA\L{,h{ .1@&&MathTypePSymbol- 2
;b 0- 2
5
&
Systems-iCompObjfObjInfoEquation Native 4_919529100F0cD<cD
FMicrosoft Equation 2.0DS EquationEquation.29q|JhJD}J
b5L, .Ole
PIC
LMETAHCompObjf1&&MathType "-MnMo0- 2
P
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qObjInfoEquation Native ,_917192732FmcD`cDOle
zJDJ0{J
2PL Q .1
&
1&MathTypeP "-W W
0m- 2
N PIC
LMETACompObjfObjInfo2
N 2
} P 0- 2
@DtJ 2
4L 2
@tJ 2
4kM 0- 2
4u(\ 2
4])\ 2
4(\ 2
4])\Symbol- 2
q= 2
U
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qP|JhJD}J
Nt(L)
=Nt(M)
PLN|N .Equation Native l_89757123058FCcDhcDOle
PIC
LMETACompObjfObjInfoEquation Native @1`&
&MathTypeP0- 2
N 0- 2
FtJ 2
L 0- 2
w(\ 2
g)\
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q$|JhJD}J
Nt(L) L0| .1`&@
&MathTypeP0- _898089814*F]}cD2 cDOle
PIC
LMETA2
%N 0- 2
RtJ 2
M 0- 2
(\ 2
)\
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q$|JhJD}J
Nt(M) CompObjfObjInfoEquation Native @_919585208Fʂ
cDcDOle
PIC
LMETAXCompObjfL{h{ .1@&`&MathTypeP0~- 2
P 0~- 2
i8
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qzJDJ0{J
PiL]
{hObjInfoEquation Native 4_919585682FB
cDETcDOle
PIC
LMETACompObjfObjInfo]
{ r .1@ &&MathTypePSymbol- 2
/e 2
Vb 0~- 2
i8 2
i8 2
i80~- 2
P 2
h 2
zSymbol- 2
= 2
-0~- 2
-( 2
) 2
1 0~- 2
+ 3
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q\zJ<J8{J
ei
=(1-Equation Native x_921396241CFcDcDOle
PIC
LPi
)b3
hi
zL{,h{ .1@&&MathTypePSymbol- 2
;b 0~- METAhCompObjfObjInfoEquation Native 4
!#$%&'()*+,-./0123469:;<?ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]_bcdefghkmnopqrstuvwxyz{|}~2
3
&
Systems-+
FMicrosoft Equation 2.0DS EquationEquation.29qzJDJ0{J
b3L{h_919586036bF0DcD7cDOle
PIC
LMETA X{ .1@&`&MathTypeP0~- 2
h 0~- 2
i8
&
Systems-
FMicrosoft Equation 2.0DS EqCompObjfObjInfoEquation Native 4_921396381FcD@'cDuationEquation.29qzJDJ0{J
hiL .1``&$ Ole
PIC
LMETACompObjf&MathType 0- 2
4z
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qzJDJ0{J
zObjInfoEquation Native (_919587215F@T5cDpcDOle
PIC
LMETA"CompObj5fObjInfo7L@q
8@q M .1&k&MathType "-``t`^ `Symbol- 2
/e 2
e 2
bSymbol- 2
t= 2
= 2
w-Symbol- 2
2
M 2
v 2
M
0~- 2
P 2
MP 2
BP 2
P 2
yh 2
z 2
P 0~- 2
i8 2
i8 2
$i8 2
S:i8 2
ki8 2
/i8 2
i8 2
Ni8 2
$
i8 2
Si8 2
ki80~- 2
( 2
E) 2
1 0~- 2
3
&
Systems-B#
FMicrosoft Equation 2.0DS EquationEquation.29qEquation Native 8_920665466FD0cDpkcDOle
=PIC
>LzJ<J8{J
e=Pi
eii
Pii
=Pi
(1-Pi
)b3
hi
zi
Pii
L! dMETA@TCompObj^fObjInfo`Equation Native a! .1&@H&MathType "-8w8rr "-`W` ``e
`
`R`<`Q-rr-``=Symbol- 2
/e 2
b Symbol- 2
8b{ 2
Yub{ 2
k
b{ 0~- 2
i8 2
Ei8 2
$i8 2
$ i8 2
Li8 2
$Fi8 2
h 2
zp 2
YCh 2
Yzp 2
kh 2
kzp0~-
2
gdP 2
P
2
dz 2
>z
2
dP
2
0dz 2
z 2
P 2
qh 2
e 2
e 2
z 2
,le`0~- 2
;|i( 2
i( 2
Ni(Symbol- 2
= 2
= 2
= 2
+ 2
,V+ Symbol- 2
-{ 2
Y-{ 2
k~-{ 0~- 2
$3 2
n2`0~- 2
;3\ 2
3\ 2
3\0~- 2
1 2
1 2
,]1 2
8( 2
) 2
,( 2
,)
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qרzJLJ({J
ei
=dPi
Pi
dzz=dPi
dzzPi
=b3
hi
e-b3hiz
(1+e-b3hiz
)2
z1(1+e-b3hiz
)Ld
d .1` &&MathType "-``Symbol- 2
<= _920666881#F;cDIcDOle
iPIC
jLMETAl2
w+ 2
= 2
- Symbol- 2
-{ 2
Y-{0~- 2
e 2
e 2
F h 2
z 2
P 0~- 2
h 2
zp 2
Yh 2
Yzp 2
i8 2
i8 2
i8`0~- 2
Eqi( 2
oi(0~- 2
h 2
z Symbol- 2
-b{ 2
Y+b{Symbol- 2
b 2
Zb`0~- 2
E3\ 2
3\0~- 2
~1 2
1 0~- 2
3 2
830~- 2
1
( 2
)
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qzJ<J8{J
=e-b3hiz
1+e-b3CompObjfObjInfoEquation Native _921399176_Q FP|cD: cDhiz
b3
hi
z=(1-Pi
)b3 e`h
i e`zL{,h{ .1@&Ole
PIC
LMETA\CompObj
f&MathTypePSymbol- 2
;b 0~- 2
3
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qzJDJ0{J
b3ObjInfo
Equation Native 4_920665465F!cD _"cDOle
L{h{ .1@&`&MathTypeP0~- 2
h 0~- 2
i8
&
SystemPIC
LMETAXCompObjfObjInfos-
FMicrosoft Equation 2.0DS EquationEquation.29qzJDJ0{J
hiLEquation Native 4_921399063F6#cDpD$cDOle
PIC
LMETACompObjfObjInfoEquation Native ( .1``&$ &MathType 0- 2
4z
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qzJDJ0{J
zL"
"
.1@&`&MathType "- (`` "-_920665463Fx%cD&cDOle
PIC
LMETA<H212
- h
R d(
d0
P-XZ
X- :
-\
T\
- 4 -R+- KSymbol- 2
/eSymbol- 2
t= 2
= 2
= 2
Z= Symbol- 2
={ 2
={ 2
j={ 2
={Symbol- 2
r
2
l 2
2
0~- 2
d 2
P 2
uP
2
!dz 2
z 2
Bd 2
\P
2
dz 2
:
z 2
P
2
2dP 2
n
2
dz 2
z
2
,dP 2
n
2
ldP 2
zz
2
Bdz
2
dP 0~- 2
i8 2
i8 2
n 2
i8 2
i8 2
n 2
ni8 2
i8 2
Ln 2
i8 2
i8 2
hn 0~- 2
y1 2
s1 2
1 2
1
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qאzJDJ0{J
e=dPPdzz=dPdzzP=dPii=1n
ndzzdPCompObj!fObjInfo"Equation Native _920666110%F ž'cD*cDii=1n
n=dPii=1n
zdzdPii=1n
L," _d," .1&TOle
PIC
$'LMETACompObj&(f"
! $#B&%'()*+,-./0132547698:;<=>?@ACEDFGIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz}~&MathType "-8?888 "-``
```Z`Symbol- 2
== 2
= 2
= Symbol- 2
<={ 2
`={ 2
$={ 2
`*
={ 2
$={ 2
`}={Symbol- 2
2
B( 2
2
B 2
r 2
BSymbol- 2
s-0~- 2
nP
2
/dP 2
P 2
z
2
dz 2
P 2
P 2
\P 2
>P 2
P 2
lh 2
z 2
P 0~- 2
[i8 2
i8 2
$i8 2
i8 2
n 2
Hi8 2
`Xi8 2
Yn 2
i8 2
i8 2
$,i8 2
xn 2
HIi8 2
`i8 2
Y
n 2
+i8 2
i8 2
Ai8 2
$i8 2
n 2
Hi8 2
`i8 2
Y_n 0~- 2
1 2
`/1 2
$
1 2
`
1 2
3 2
$y1 2
`10~- 2
{1Symbol- 2
(e 2
b0~- 2
( 2
A)
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qObjInfo)Equation Native _898335547,F@+cDr,cDOle
"')*+,-/46789:;<=?BEGHIJKLMNOPQRSTUVX[\_abcdeglnopqrty{|}~zJDJ0{J
e`=
Pi
dPi
Pi
zdzi=1n
Pii=1n
e`=
Pi
eii=1n
Pii=1n
e`=
Pi
(1-Pi
)b3
hi
zi=1n
Pii=1n
L| .1``&
&MathTypeP0- 2
N 0- PIC
+.
LMETACompObj-/fObjInfo02
HtJ 2
M 0- 2
y(\ 2
)\
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q$|J`J<}J
Nt(M) Equation Native @_8978339693F0.cD).cDOle
PIC
25LL| .1``&
&MathTypeP0- 2
N 0- 2
HtJ 2
M 0- 2
y(\ 2
)\
&
METACompObj46!fObjInfo7#Equation Native $@ Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q$|JhJD}J
Nt(M) LE{lh_8978340431?:F]/cD0cDOle
%PIC
9<&LMETA(`E{ .1@&&MathTypeP0-2
&POP/ 0- 2
3tJ
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qCompObj;=.fObjInfo>0Equation Native 1<_897834047AFP1cDP03cD }JJh}J
POPtLq|q .1` &
&MathTypeP0- Ole
2PIC
@C3LMETA5CompObjBD>f2
Y 0- 2
tJ 2
M Symbol- 2
O-{ 0- 2
1 2
1(\ 2
)\
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qObjInfoE@Equation Native AD_898335405HFgE4cD>5cDOle
C(}JJh}J
Yt-1(M)LIx|I .1`&
PIC
GJDLMETAFCompObjIKWfObjInfoLY&MathTypeP0- 2
N 2
POP/ 2
IY 2
Y 0- 2
M 2
M 2
5M 0-
2
T94
2
94
2
A93
2
930- 2
C02
406 0- 2
y(\ 2
)\ 2
(\ 2
)\ 2
(\ 2
+)\0- 2
( 2
) 2
+.Symbol- 2
= 2
k
- 2
" 2
+
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qޜ}JJh}J
N94(M)
=(POP94
-Y93(M)
)0.406+Y93(M)L|Equation Native Z_920662090OFP6cDx_7cDOle
]PIC
NQ^LMETA`pCompObjPRffObjInfoShEquation Native i@ .1`&
&MathTypeP0- 2
V 0- 2
i82
"(L)\\
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q$|JhJD}J
Vi(L) LN|N .1`&
&MathTypeP0- _920662089[MVFХ8cD9cDOle
jPIC
UXkLMETAmp2
V 0- 2
i82
$(M)\\
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29q$|JhJD}J
Vi(M) CompObjWYsfObjInfoZuEquation Native v@_920662088]Fcy:cDcD>?cDOle
PIC
cfLMETA\CompObjegfObjInfoh&MathTypePSymbol- 2
;b 0~- 2
1
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qzJDJ0{J
b1Equation Native 4_920662086kFEAcDjAcDOle
PIC
jmLL{,h{ .1@&&MathTypePSymbol- 2
;b 0~- 2
2
&
SystMETA\CompObjlnfObjInfooEquation Native 4ems-
FMicrosoft Equation 2.0DS EquationEquation.29qzJDJ0{J
b2L{,h_920662085~irF0.CcD DcDOle
PIC
qtLMETA\{ .1@&&MathTypePSymbol- 2
;b 0~- 2
3
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qCompObjsufObjInfovEquation Native 4_920662084yF@uHFcD@yGcDzJDJ0{J
b3L{,h{ .1@&&MathTypePSymbol- Ole
PIC
x{LMETA\CompObjz|f2
;b 0~- 2
4
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qzJDJ0{J
b4ObjInfo}Equation Native 4_920662083wFHcDyIcDOle
PIC
LMETA\CompObjfObjInfoL{,h{ .1@&&MathTypePSymbol- 2
;b 0~- 2
5
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qzJDJ0{J
b5L{,h{ .Equation Native 4_898336401FJcDKKcDOle
PIC
LMETAXCompObjfObjInfoEquation Native 41@&&MathTypeP0- 2
C 0- 2
3i8
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qf}JJh}J
CiL{<h{ l .1@`& &MathTypeP0-
2
(( 2
9, 2
, 2
;_897563514F`jLcD@qMcDOle
PIC
LMETA) 2
. 2
)2
w6002
0002
7322
y000 2
0
2
01Symbol- 2
+ 2
2
+0- 2
C 0- 2
i8
&
Systems-CompObjfObjInfoEquation Native _898114098FFNcDZOcD
FMicrosoft Equation 2.0DS EquationEquation.29qfd|JJ`}J
((600,000+732,000)0.01+Ci
)L{hOle
PIC
LMETAXCompObjf{ .1@&`&MathTypeP0- 2
#h 0- 2
i8
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qObjInfoEquation Native 4_898114097F``QcD 2
ScDOle
}JJh}J
hiL{h{ .1@&&MathTypeP0- PIC
LMETA<CompObjfObjInfo
"#$%&'(*2
( 2
B, 2
, 2
7
. 2
, 2
, 2
5) 2
12
2002
000 2
L 0
2
01
2
162
000 2
;12
c600Symbol- 2
2
+ 2
+ 2
n0- 2
h 0- 2
i8
&
Systems-
FMicrosoft Equation 2.0DS EquationEquation.29qt}JJh}J
(1,200,0000.01+16,000+1,600hi
)Equation Native _898195279F1UcDPvXcDOle
PIC
LL/\
FũμƮ ƮBiff5Excel.Chart.59qOh+'0@H`xBook
HdMETA|CompObjeObjInfo
sB=<$$8X11111111"\"#,##0;\-"\"#,##0"\"#,##0;[Red]\-"\"#,##0"\"#,##0.00;\-"\"#,##0.00!"\"#,##0.00;[Red]\-"\"#,##0.003*0_-"\"* #,##0_-;\-"\"* #,##0_-;_-"\"* "-"_-;_-@_-*)'_-* #,##0_-;\-* #,##0_-;_-* "-"_-;_-@_-;,8_-"\"* #,##0.00_-;\-"\"* #,##0.00_-;_-"\"* "-"??_-;_-@_-2+/_-* #,##0.00_-;\-* #,##0.00_-;_-* "-"??_-;_-@_-\$#,##0_);\(\$#,##0\)\$#,##0_);[Red]\(\$#,##0\)\$#,##0.00_);\(\$#,##0.00\)# \$#,##0.00_);[Red]\(\$#,##0.00\)0.00.000 + ) , * 8 ``i̜̙3f3333f3ffff333ff333f33f33BBBSheet17
Sheet1
WSheet2
XSheet3
XSheet4
YSheet5
ZSheet6
[Sheet7
\Sheet8
g]Sheet9M^Sheet103_Sheet11`Sheet12`Sheet13aSheet14bSheet15cSheet16
s
dMbP?_*+% RR"R??U
>
s
dMbP?_*+% RR&A&P ʃ&~?'~?(~?)~?MHP LaserJet(Korea)??B t1\j9 W0%X%T@X" K,`?`?U}3 }f}}
(
b
b
b
b
b
b
W
P
GP@
b@
b
b
bb
bCellular
M94=20 ̻M94=20 ̻-64 Ͼ
LEO
YEAR
St
Yt Fitted St Fitted YtN94=(M94-Y93)*0.406+Y93N94=(M94-Y93)*0.406+Y93Mt=20-Mt=20-64
St
YtNt=(Mt-Yt-1)*0.47+Yt-1
~
until 1993'increase with population growth'increase with population growth
until 1993
@Ĥ@Ĥ@?Qr1@&)t^c:?}"?DDDDQr1@& DD$ӡi3bADDD$(SKaA
DDD~
)tADD D @D@M@?"X@)t^c:?}"?DDDDz@ DD$facADDD$WcaADDD"QuA @DТ@@?T@)t^c:?}"?DDDDpR@ DD$
vbcADDD$Bү+t1bADDD~
b)CuAD DD @DȨ@@?arr@)t^c:?}"?DDDD0y&@ DD$*~^dADDD$sbADDD~
y
)evAD DD @D@@@?[Я!@)t^c:?}"?DDDDt @ DD$Gr>[dADDD$[_cADDD~
4)vAD DD @ D@@d@?Ѻ&@)t^c:?}"?DDDDhtR@ DD$dfW#eAD DD $cAD DD ~
"n\)ywAD DD @
D @P@? p<@ )t^c:?}"?DD D D A@
DD $ ZTGeAD
D D
$ s(IcAD
D D
bb
@D
@IA?
ygUB\@
)t^c:?}"?D D
D
D
ޙo8[A D D
$
.fA DD
D$
zxdA DD
D
zA%
@D
`@A?@)t^c:?}"?D
DDD
*/AA D
D$oXADGz?@@D
?\4;A)t^c:?}"?DDDD{_UgA DD$}
3jADDD$;cgADDDnb"! ?
w($@)t^c:?}"?DDDD2@
DD
$⭧XADGz?D@D
?Y<g0A)t^c:?}"?DDDDwbiA DD$>kW͏jADDD$>}gADDD~
Y)|ADDD ?
,
1@)t^c:?}"?DDDD/XyJd@
DD
$IaXADGz?H@D
?&+
A)t^c:?}"?DDDD ^4acojA DD$.jADDD$@l6hADDD~
D)}ADDD ?
NU]@)t^c:?}"?DDDD @
DD
$iOYADGz?L@D
?'A)t^c:?}"?DDDDHW kA DD$jUIkADDD$B[shADDD~
b/6)_}ADDD ?
qY}/@)t^c:?}"?DDDD&@
DD
$J/7YADGz?P@D
?^A)t^c:?}"?DDDD:blwkA DD$݂=kADDD$EJ;(shADDD~
BR)}ADDD ?
(*c@)t^c:?}"?DDDDy6A
DD
$}+jYADGz?T@D
?u%A)t^c:?}"?DDDDTwkA DD$PDzlADDD$G9qUhADDD"n| ?
*~@)t^c:?}"?DDDDTqA
DD
$TZADGz?X@D
?b'XA)t^c:?}"?DDDD>t#lA DD$X`AlADDD$r]iADDD)A
DDD)~A&DDD ?
ǂmXA)t^c:?}"?DDDD;A
DD
$_ZADGz?\@D
?&koA)t^c:?}"?DDDDMelA DD$'lADDD$ԅt0iADDD)H'ADDD)7C~ADDD ?
A;A)t^c:?}"?DDDD#'A
DD
$n0b!ZADGz?`@D
?/:v@)t^c:?}"?DDDD_(lA DD$ $hclADDD$,vjRiADDD)LADDD)k~ADDD ?
nYA)t^c:?}"?DDDD3A
DD
$2,[ADGz?d@D
?pMye@)t^c:?}"?DDDDlA DD$nlADDD$aw[tiADDD)@rADDD),~ADDD ?
KC'A)t^c:?}"?DDDDt6,?A
DD
$bUǤ?[ADGz?h@D
?
TL@)t^c:?}"?DDDDr=+q"mA DD$Ug;mADDD$xyΕiADDDB) ?
zYI-0A)t^c:?}"?DDDD#GA
DD
$yfz[ADGz?l@D
?&]E@)t^c:?}"?DDDD1ZmA DD$poqmADDD$pQӭiADDD)AD"DD)~AD"DD ?
53A)t^c:?}"?DDDDBPA
DD
$E
0[ADGz?p@ D
?CZ
b@)t^c:?}"?DDDDmmA DD$FvmADDD$hiA DDD)0فA D"DD)^~A D"DD ?
St%Z3A)t^c:?}"?DDDD$aUA
DD
$2[A DGz?D)/l
;;;;,O
b !
"
#
b $
b%
b&
'
t@!D
? /@ )t^c:?}"?DD D D gmA! DD $ q1~mA DD D$ r`iA!DD D) A!D"DD) 9A!D"DD ?
],A )t^c:?}"?DD D D jZ1YA!
DD
$ Xl\A!D Gz?!x@"D
!?!4-b@!)t^c:?}"?D D!D!D !6u+mA" D D!$!YnA!D D!D $!`X5iA"D D!D )!:AD"DD )!0AD"DD ! ?!
):A!)t^c:?}"?D D!D!D !/Y-[A"
D D!
$! J~E\A"D!Gz?"|@#D!
"?"mc@")t^c:?}"?D!D"D"D!"%4nA# D!D"$"A&GInA"D!D"D!$"NP
jA#D!D"D!"_ "<" ?"
^[@dA")t^c:?}"?D!D"D"D!"
\A#
D!D"
$":Vx\A#D"Gz?#@$D"
#?#%f1Yy@#)t^c:?}"?D"D#D#D"#_PAcnA$ D"D#$#uXunA#D"D#D"$#'~jA$D"D#D")#CUAD'D"D")#TAD'D"D"# ?#
}>(jA'D%D&D%)&ͤA%D'D"D%)&lA%D'D"D%& ?&
麧w@&)t^c:?}"?D%D&D&D%&T
]A'
D%D&
$&~]A'D&Gz?'@D&
'?'Lf)n @')t^c:?}"?D&D'D'D&'"zoA
D&D'$'po%oA'D&D'D&$'z.jAD&D'D&'_ b' ?'
@
@')t^c:?}"?D&D'D'D&'E6]A D&D'
$'g`G]AD'Gz?]>+: @A@Pqn U sMHP LaserJet(Korea)7 t1\j9 W0%X%T@X" d??333QQ;;qQ;WqE43QQ;sqQ;qE43QQ;qQ;q3
4E43QQ;qQ;q3
3
34E43QQ;rQ;3
4E4D
FA3OJ
n343*!
4523
43"
$%3O&Q4444
6
]>+>5@A@,j U s&A&P ʃMHP LaserJet(Korea) %X%T@X" d??3Z33QQ;[oQ;E4D
FA<A 3Oz
3 43*!
4%
#3O) &Q
Year'4%8;c 3O2&Q
'$The number of cumulative subscribers'4523
43"
$%3O&Q4444
>
+000"
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
s
dMbP?_*+% RR&A&P ʃ"Rf??crU
>
/\
J""")))UUUMMMBBB999|PP3f333f3333f3ffffff3f̙3ff333f333333333f33333333f33f3ff3f3f3f3333f33̙33333f333333f3333f3ffffff3f33ff3f3f3f3fff3ffffffffff3ffff̙fff3fffff3fff333f3f3ff3ff33f̙̙3̙ff̙̙̙3f̙3f333f3333f3ffffff3f̙3f3f3f333f3333f3ffffff3f̙3f3ffffffffff!___www45' E-0E- Systems-'- J-'- J- "-
$7
7
"- "---'--- 7
7
::7
7
7
AA7
7
---'--- J "-
7
7
"--::AA7
mmmmmmm[m[I mI 7
m7
---'--- )!!---'--- 8
---'--- O
- "-
sY,m[[yI 7
-$mms$^ss^Y$DYn}YD,$,Ak,m$mmXm[[$[Fp[[pF[[FyI $I d^ yI 4 yI d7
$7
L
7
"
7
---'--- 8
---'--- J---'--- J---'--- J
2
`6062
^50000666662
(1000006666662
g(1500006666662
(2000006666662
(2500006666662
m(300000666666---'--- J---'--- J
- 2
~1998-- 2
l1999-- 2
Z2000-- 2
H2001-- 2
52002-- 2
#2003-- 2
2004-- 2
2005----'--- J---'--- Fg
2
Year:66---'--- J----'---
0-2
The number of cumulative-0-
2
subscribers----'--- J---'--- J-- -'J ' -4'SummaryInformation(DocumentSummaryInformation81Table+ SummaryInformation(ѱбѱбMMicrosoft Excel@w@wjZ[՜.+,0\8@`h
p;濵/ Sheet17Sheet1Sheet2Sheet3Sheet4Sheet5Sheet6Sheet7Sheet8Sheet9Sheet10Sheet11Sheet12Sheet13Sheet14Sheet15Sheet16 ũƮOh+'0 0<
Xdp
|(Forecasting ATM Service Demand in Koreaore한국과학기술원e Normal.dotSunTrust Bank Inc2nTMicrosoft Word 8.0@F#@F.q@:bD@:bDhF՜.+,D՜.+,h$hp
SunTrust Banks, Inc. $V1(Forecasting ATM Service Demand in KoreaTitle 6>
_PID_GUIDAN{EC0C5B69DocumentSummaryInformation8!CompObj)j-57A0-11D6-ADC0-0002A5532CEA}
FMicrosoft Word Document
MSWordDocWord.Document.89q
[>@>Normal$1$dhH$4$9DmH nH<A@<Default Paragraph Font, @,Footer
8!&)@&Page Number,@,Header
8!U:
F
t"'06|=iAJGHJ,MUUI-
VO
%(g-9IN&QwSXY.24567:<>AEc%;{MeO+QR(TUYY/138;=?@BDFHdUY09C*,Vjl.BD !1!3!@!T!V!d!x!z!!!!="Q"S"X"l"n"#$$9$M$O$U$i$k$$$$$$$''''''''''( (P)d)f)p)))90M0O0000000000n4445
66 64666i6}66666788{IIIIIIIIIIIJJJJJJJJJJfKzK|KKKKL#L%L]LqLsLM&M(MMMMMMMMMNN-N/NNNNcOwOyOOOOP$P&PRRRSSSTThTjTTU
UUUUU::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::!!x~$07DVmorsy.EHKLU!4!6i66666678
8
888dBkBpBuBBBBBQCVC^CdCDDKDDDgEnEEEFFFFRG\GGGGGHHHHI"I{IIIIIIIIIIIJJJJJJJJJfK}K~KKKKL&L(L+L,L2L]LtLvLyLzLLM)M,M1M2M3MMMMMMMMMMMNNN0N1N*CJOJQJo(. @.W.7\7\DR'QQQ @W.P 4@MW56>*CJOJQJo(@\\SUNTRUST_TREE\ga030p40_aqa_hp8150_q4.aqalr.303pea.atl.ga.stiNe00:winspoolHP LaserJet 8150 PS\\SUNTRUST_TREE\ga030p40_aqa_hp
odXXLetterPRIV''''\KhC1\\SUNTRUST_TREE\ga030p40_aqa_hp
odXXLetterPRIV''''\KhC18U@@GTimes New Roman5Symbol3&Arial71Courier7
pд5 pд/???ؽд;Wingdings 0Rhszhszh1thF$f(VY!!%),.:;?]} 0 00
0002 3 ! =]([\{ 00
0000;[pxxV|/IR'Forecasting ATM Service Demand in Korea\mY0 SunTrust Bank Inc