Earticle

본 연구는 2003년 1사분기부터 2016년 2사분기 까지 인천국제공항에서 미주노선을 통하여 미주 내 공항에 도착하는 항공화물의 시계열 자료를 통하여 SARIMA 모형을 활용하여 항공화물 수요예측을 시행하였다. 또한 SARIMA 모형을 활용하여 만들어진 수요예측 모형과 기존 연구에 주로 활용되어졌던 ARIMA 모형을 활용하여 만들어진 수요예측 모형과 비교분석함으로써, 주기적인 특성 및 계절성을 가진 시계열 자료에 대한 SARIMA 모형의 상대적으로 우수한 예측 정확성을 입증하였다. 기존의 항공 관련 연구는 주로 여객에 관한 연구가 상대적으로 많았다. 또한 화물과 관련된 연구에서도 특정노선이 아닌 공항이나 전체에 대한 연구가 대부분이었다. 이러한 상황에서, SARIMA 모형을 활용하여 미주지역이라는 특정 노선에 대한 항공화물의 수요를 예측한 본 연구는 큰 의의가 있다고 생각된다.

For forecasting air cargo demand from Incheon National Airport to all of airports in the United States (US), this study employed the Seasonal Autoregressive Integrated Moving Average (SARIMA) method and the time-series data collected from the first quarter of 2003 to the second quarter of 2016. By comparing the SARIMA method against the ARIMA method, it was found that the SARIMA method performs well, relatively with time series data highlighting seasonal periodic characteristics. While existing previous research was generally focused on the air passenger and the air cargo as a whole rather than specific air routes, this study emphasized on a specific air cargo demand to the US route. The meaningful findings would support the future research.

Semiconductor industry has been taking the advantage of improvements in process technology in order to maintain reduced device geometries and stringent performance specifications. This results in semiconductor manufacturing processes became hundreds in sequence, it is continuously expected to be increased. This may in turn reduce the yield. With a large amount of investment at stake, this motivates tighter process control and fault diagnosis. The continuous improvement in semiconductor industry demands advancements in process control and monitoring to the same degree. Any fault in the process must be detected and classified with a high degree of precision, and it is desired to be diagnosed if possible. The detected abnormality in the system is then classified to locate the source of the variation. The performance of a fault detection system is directly reflected in the yield. Therefore a highly capable fault detection system is always desirable. In this research, time series modeling of the data from an etch equipment has been investigated for the ultimate purpose of fault diagnosis. The tool data consisted of number of different parameters each being recorded at fixed time points. As the data had been collected for a number of runs, it was not synchronized due to variable delays and offsets in data acquisition system and networks. The data was then synchronized using a variant of Dynamic Time Warping (DTW) algorithm. The AutoRegressive Integrated Moving Average (ARIMA) model was then applied on the synchronized data. The ARIMA model combines both the Autoregressive model and the Moving Average model to relate the present value of the time series to its past values. As the new values of parameters are received from the equipment, the model uses them and the previous ones to provide predictions of one step ahead for each parameter. The statistical comparison of these predictions with the actual values, gives us the each parameter's probability of fault, at each time point and (once a run gets finished) for each run. This work will be extended by applying a suitable probability generating function and combining the probabilities of different parameters using Dempster-Shafer Theory (DST). DST provides a way to combine evidence that is available from different sources and gives a joint degree of belief in a hypothesis. This will give us a combined belief of fault in the process with a high precision.

예측의 정확성은 비용의 감소나 고객서비스의 제고를 위해 필수적으로 선행되어야 하기에 현재까지도 많은 연구자들에 의해 연구되고 있는 분야이다. 본 연구에서는 국내 항만의 컨테이너 물동량 예측에 있어 대표적인 비선형예측모형인 인공신경망모형과 ARIMA모형에 대한 비교연구를 수행하는데 목적을 두었고, 컨테이너 물동량 예측력 제고를 위해 ARIMA모형과 인공신경망(ANN)모형을 결합한 하이브리드모형을 사용해 다른 모형들과 예측성과를 비교하고자 한다. 특히 인공신경망모형의 네트워크 구조 설계에 부분에 있어 방대하며 복잡한 탐색공간에서도 전역해 찾기에 효과적인 기법으로 알려져 있는 유전알고리즘을 사용함과 동시에 인공신경망의 대표적인 모형으로 알려진 다층 퍼셉트론(MLP)뿐만 아니라 시간지연네트워크(TDNN)를 사용해 예측성과를 비교하였다. 그 결과 ANN모형과 하이브리드모형이 ARIMA모형보다 더 뛰어난 예측성과를 보이는 것으로 나왔다.

The accuracy of forecasting is remarkably important to reduce total cost or to increase customer services, so it has been studied by many researchers. In this paper, the artificial neural network (ANN), one of the most popular nonlinear forecasting methods, is compared with autoregressive integrated moving average(ARIMA) model through performing a prediction of container traffic. It uses a hybrid methodology that combines both the linear ARIAM and the nonlinear ANN model to improve forecasting performance. Also, it compares the methodology with other models in performance for prediction. In designing network structure, this work specially applies the genetic algorithm which is known as the effectively optimal algorithm in the huge and complex sample space. It includes the time delayed neural network (TDNN) as well as multi-layer perceptron (MLP) which is the most popular neural network model. Experimental results indicate that both ANN and Hybrid models outperform ARIMA model.

In this paper we consider the problem of testing for structural changes in ARIMA models based on a cusum test. In particular, the proposed test procedure is applicable to testing for a change of the status of time series from stationarity to nonstationarity or vice versa. The idea is to transform the time series via differencing to make stationary time series. We propose a graphical method to identify the correct order of differencing.

본 논문에서는 단기 수요예측방법으로 ARIMA모형을 소개하며, 실전에서 이 모형을 설정하기 위해서 필요한 경험적인 전략에 초점을 두었다. 이 실전전략을 통해 프로세서 내에 복잡하게 얽혀있는 메커니즘을 단순용이하게 설명 및 규명할 수 있으며, 더 나아가 기업의 판매계획, 생산계획, 인사계획, 재무계획 등에 필요한 의사결정을 하는데 도움이 될 수 있다.

In this study, ARIMA model is introduced, which is one of the widely used short-term demand forecasting methods and empirical strategies which can be applied to the real fields are focused. We can easily and simply explain and/or diagnose the mechanism of process and furthermore obtain the important information to make decisions by applying these 31 strategies.

This research explores how imported automobile companies can develop their strategies to improve the outcome of their recalls. For this, the researchers analyzed patterns of recall demand, classified recall types based on the demand patterns and examined response strategies, considering plans on how to procure parts and induce customers to visit workshops, recall execution capacity and costs. As a result, recalls are classified into four types: U-type, reverse U-type, L- type and reverse L-type. Also, as determinants of the types, the following factors are further categorized into four types and 12 sub-types of recalls: the height of maximum demand, which indicates the volatility of recall demand; the number of peaks, which are the patterns of demand variations; and the tail length of the demand curve, which indicates the speed of recalls. The classification resulted in the following: L-type, or customer-driven recall, is the most common type of recalls, taking up 25 out of the total 36 cases, followed by five U-type, four reverse L-type, and two reverse U-type cases. Prior studies show that the types of recalls are determined by factors influencing recall execution rates: severity, the number of cars to be recalled, recall execution rate, government policies, time since model launch, and recall costs, etc. As a component demand forecast model for automobile recalls, this study estimated the ARIMA model. ARIMA models were shown in three models: ARIMA (1,0,0), ARIMA (0,0,1) and ARIMA (0,0,0). These all three ARIMA models appear to be significant for all recall patterns, indicating that the ARIMA model is very valid as a predictive model for car recall patterns. Based on the classification of recall types, we drew some strategic implications for recall response according to types of recalls. The conclusion section of this research suggests the implications for several aspects: how to improve the recall outcome (execution rate), customer satisfaction, brand image, recall costs, and response to the regulatory authority.

We consider fractional Brownian motion and FARIMA process with Gaussian innovations and show that the suitably scaled distributions of the FARIMA processes converge to fractional Brownian motion in the sense of finite dimensional distributions. We figure out ACF function and estimate the self-similarity parameter H of FARIMA(0, d, 0) by using R/S method. Finally, we display power spectrum density of FARIMA process.

We study the weak convergence of various models to Fractional Brownian motion. First, we consider arima process and ON/OFF source model which allows for long packet trains and long inter-train distances. Finally, we figure out power spectrum density as a Fourier transform of autocorrelation function of arima model and binary signal model.

우리나라 멸치어업에서의 1971~1992년 동안의 22년간 월별 어획량 자료를 시계열 분석하여 어획량을 분석, 예측하였다. 시계열 분석은 다른 생물학적, 해양학적, 사회 경제적인 요소가 없어도 단지 어획량 자료만으로 분석과 예측이 가능하다. 첫 20년간인 1971~1990년 사이의 월별 멸치 어획량 자료를 ARIMA 시계열 모형에 적용시켜 구한 결과는 다음과 같다. 로그 (대수) 변환시켰을 때의 ARIMA 모형: $$(1-0.381B)(1-0.027B^{12}+0.141B^{24})(1-B^1)(1-B^{12})Z_t=(1-0.968B)(1-0.727B^{12})e_t$$, Box-Cox 변환시켰을 때의 ARIMA 모형: $$(1-0.431B)(1-B^{12})Z_t=(1-0.882B^{12})e_t$$, 위의 두 모형중 Box-Cox 변환시킨 것이 로그 (대수) 변환시킨 것보다 예측오차가 적었으며, Box-Cox 변환식은 $Y'=(Y^{0.58}-1)/0.58$ 이었다. 위의 두 모형 중 후자의 모형을 이용하여 1991~1992년 사이의 월별 어획량을 예측하였다. 예측 어획량과 실제 어획량과의 월별 오차범위는 1.0~63.2% (1991년에 1.6~63.2%이고, 1992년에는 1.0~60.4%)였다. 예측 어획량이 각 연도별로 148,201M/T과 148,834M/T인데 비해, 실제 어획량은 170,293M/T, 168,234M/T이었다. 2년 동안의 총어획량에 대한 오차는 12.3%였다. 또한 스펙트럼 분석은 순환변동의 주기가 2.2개월, 6.1개월, 10.2개월, 12개월, 14.7개월에서 상대적으로 큰 성분이 있음을 나타내었다 이 순환변동 성분은 적절한 ARIMA 모형을 결정하는 데도 도움이 된다.

Forecasts of the monthly catches of anchovy in Korea were carried out by the seasonal Autoregressive Integrated Moving Average (ARIMA) model and spectral analysis. The seasonal ARIMA model is as follows: $$(1-0.431B)(1-B^{12})Z_t=(1-0.882B^{12})e_t$$ where: $Z_t=value$ at month $t;\;B^{p}$ is a backward shift operator, that is, $B^pZ_t=Z_{t-p};$ and $e_t=error$ term at month t, which is to forecast 24 months ahead the anchovy catches in Korea. The prediction error by the Box-Cox transformation on monthly anchovy catches in Korea was less than that by the logarithmic transformation. The equation of the Box-Cox transformation was $Y'=(Y^{0.58}-1)/0.58$. Forecasts of the monthly anchovy catches for $1991\~1992$, which were compared with the actual catches, had an absolute percentage error (APE) range of $1.0\~63.2\%$. Total observed annual catches in 1991 and 1992 were 170,293 M/T and 168,234 M/T respectively, while the predicted catches were 148,201 M/T and 148,834 M/T $(API\;13.0\%\;and\;11.5\%,\;respectively)$. The spectrum analysis of the monthly catches of anchovy showed some dominant fluctuations in the periods of 2.2, 6.1, 10.2 12.0 and 14.7 months. The spectrum analysis was also useful for selecting the ARIMA model.

계절변동조정에 대한 전문적인 지식이 없는 통계작성(이용)자가 계절변동조정통계를 정교히 작성하기 위해서는 계절변동조정작업을 전산화하여 체계화할 필요가 있다. 본고에서는 X-12-ARIMA에 바탕은 둔 BOK-X-12-ARIMA프로그램의 개발 내역을 설명하고 향후 개선방향을 정리하였다.

To compile seasonally-adjusted statistics for Korean economic statistics accurately. it is necessary to develop a Korean seasonal adjustment program. In this paper. the Korean seasonal adjustment program BOK-X-12-ARIMA, developed through modification of the US. Bureau of the Census's X-12-ARIT\IA, is explained in detail.

An ARIMA signal observed with measurement error is shown to have another ARIMA representation with nonlinear restrictions on parameters. For this model, the restricted Newton-Raphson estimator(RNRE) of the unit root is shown to have the same limiting distribution as the ordinary least squares estimator of the unit root in an AR(1) model tabulated by Dickey and Fuller (1979). The RNRE of parameters of the ARIMA(p,1,k) process and unit root tests base on the RNRE are developed.

In call forecasting literature, both the seasonal autoregressive integrated moving average(ARIMA) type models and seasonal linear models have been popularly suggested as competing models. However, their parallel comparison for the forecasting accuracy was not strictly investigated before. This study evaluates the accuracy of both the seasonal linear models and the seasonal ARIMA-type models when predicting intra-day call arrival rates using both real and simulated data. The seasonal linear models outperform the seasonal ARIMA-type models in both one-day-ahead and one-week-ahead call forecasting in our empirical study.

The purpose of this study is to develop the stochastic stream water quality model for the intake station of Chung-Ju city waterworks in the Han river system. This model was based on the theory of Box-Jenkins Multiplicative ARIMA(SARIMA) and the state space model to simulate changes of water qualities. Variable of water qualities included in the model are temperature and dissolved oxygen(DO). The models development were based on the data obtained from Jan. 1990 to Dec. 1997 and followed the typical procedures of the Box-Jenkins method including identification and estimation. The seasonality of DO and temperature data to formulate for the SARIMA model are conspicuous and the period of revolution was twelve months. Both models had seasonality of twelve months and were formulates as SARIMA {TEX}$(2,1,1)(1,1,1)_{12}${/TEX} for DO and temperature. The models were validated by testing normality and independency of the residuals. The prediction ability of SARIMA model and state space model were tested using the data collected from Jan. 1998 to Oct. 1999. There were good agreements between the model predictions and the field measurements. The performance of the SARIMA model and state space model were examined through comparisons between the historical and generated monthly dissolved oxygen series. The result reveal that the state space model lead to the improved accuracy.

본고는 2012년까지의 해상물동량을 예측하고 항만물류정책적 방안을 제시하는데 목적을 두었다. ARIMA 모형을 통한 분석을 위해서 1차적으로 모형을 식별하였다. 자기상관도표를 통해 물동량의 자기상관함수값이 대단히 느린 속도로 0에 접근하여 안정적이지 못한 것으로 나타났으나, 자기상관계수가 1차차분 후 시차1 이후 급격한 감소를 보임에 따라 AR(1) 과정을 갖는다는 것을 알 수 있었다. 또한 자료들이 강한 계절성을 갖는 것으로 나타남에 따라 식별단계를 거쳐 승법계절 ARIMA모형인 ARIMA(1,1,1)(1,0,1)s 모형을 도입하였다. 다음 단계로 2007년부터 2012년까지의 사전적 예측치를 살펴보았다. 그 결과 2007년 6억9,631만톤, 2008년 7억2,180만톤, 2009년 7억4,807만톤, 2010년 7억7,520만톤, 2011년 8억320만톤, 2012년 8억3,212만톤으로 매우 느리게 증가하였다. 2006년 대비 증가율로 보면 2007년 1.42%, 2009년 8.96%, 2012년 21.21%로 나타났다. 구체적으로 입하량의 경우는 2007년 0.86%에서 2012년 16.1%로 증가하며, 출하량의 경우는 2007년 2.76%에서 2012년 33.2%로 증가함을 알 수 있었다. 그리고 항만물동량 증가추세 둔화현상의 극복과 항만의 로컬 화물 창출 및 부가가치 창출 기능을 위해서 제조업의 공동화 억제, 환적화물의 지속적이고 적극적인 유치, 항만배후물류단지의 조기 개발과 다국적 기업의 유치, 한 중 물류협력 강화, 복합운송체계의 구축을 제시하였다.

The purpose of this study is to forecast the marine trading volumes using multiplicative seasonal Autoregressive Integrated Moving Average(ARIMA) model. The paper proceeds by comparing the forecasting performances of the unload volumes with those of the load volumes with Box-Jenkins ARIMA model. Also, I present the predicted values based on the ARIMA model. The result shows that the trading volumes increase very slowly.

본 연구에서는 미국․캐나다를 비롯하여 세계적으로 가장 널리 사용되는 X-12-ARIMA 유형의 계절조정방법과 유럽에서 많이 사용되고 있는 TRAMO-SEATS 방법을 중심으로 각 계절조정법의 특징에 대한 이론적․실증적 비교분석을 시도하였다. 특히 한국의 주요 거시경제지표에 대하여 X-12-ARIMA 방법과 TRAMO-SEATS 방법을 각각 적용하여 계절조정계열을 산출하고, 멱등성․안정성 등 계절조정의 적합성 평가기준을 토대로 두 방법의 유용성을 비교하였다. 본 연구의 분석 결과에 의하면 두 방법 모두 안정성이 우수한 것으로 나타나 정도 높은 계절조정계열을 산출하는 것으로 평가되었다. 두 방법 사이의 상대적 비교에서는 대상 자료에 따라 약간 다른 결과를 보이고 있기는 하지만 전체적으로 X-12-ARIMA 방법이 다소 우월한 것으로 나타났다. 그러나 서로 다른 조건과 통계적 기준으로 계절조정을 시행하는 여러 가지 계절조정방법을 실증적인 기준으로 비교한 본 연구의 결과를 해석하는 데는 신중을 기해야 한다. 따라서 본고의 분석결과를 반영하여 X-12-ARIMA 방법만 사용하는 것보다는 두 방법을 같이 병행하면서 각각을 한국 경제시계열 자료에 적합하도록 조정하는 방안을 강구해야 하며, 이에 관한 지속적인 연구를 통해 신호추출법에 기초한 TRAMO-SEATS 방법의 이론적 정합성과 X-12-ARIMA 방법의 실증적 적합성을 결합시킬 수 있는 새로운 차원의 계절조정방법을 계속 모색해야 할 것이다.

The forecasting of container volume which is the basis of port logistics facilities expansion has a great influence on development of an port. Based on this importance, various previous studies have presented methodology on container volume forecasting. The results of many previous studies pointed out the limitations of future forecasting based on past container volume and emphasized that more various factors should be considered to compensate this. Taking notice of this point, this study forecasted future container volume by using ARIMA model, time series analysis and System Dynamics (SD) method, a dynamic analysis technique and performed the comparative review with the forecast of the Ministry of Land, Transport and Maritime affairs. Recently with rapid changes in economic and social environment, the non-linear change tendency for forecasting container traffic is presented as a new alternative to the country.

Recently, the oil suppliers who supply the tax-exempt oil to the fishery are confronted with big trouble in their supply and demand system due to the unstable global oil prices. We applied the seasonal ARIMA(SARIMA) model to the low-sulfur and high-sulfur crude oil which are in great request and developed forecasting systems for them. Since there are many parameters in SARIMA, it is difficult to estimate the optimal parameters, but it is overcome by using simulation looping program. In conclusion, we found that the obvious seasonality in demand of low-sulfur and these demands are tending downwards gradually.

BOK-X12-ARIMA has been developed by the Bank of Korea in order to accomodate special features such as lunar effect, labor day and election effect which are intrinsic in Korean seasonal time series. Irregular component resulting from BOK-X12-ARIMA is usually treated as white noise time series. If this shows dependency structure, it may be advisable to incorporate dependency in irregular component into prediction. This article illustrates how to refine point forecast using dependency structure in irregular component.

The X-12-ARIMA program was utilized on the analysis of the time series trend on 76 Korean industrial activities data in order to ensure that the trading day effect adjustment as well as the seasonal effect adjustment is needed to extract the fundamental trend-cycle factors from various economic time series data. The trading day effect is strongly correlated with the activity of production and shipping but not with the activity of inventory. Furthermore, the industrial activities were classified with respect to the sensitivity on the tranding day effect.