• 대한산업공학회지
• /
• 제34권2호
• /
• pp.122-139
• /
• 2008

With the proliferation of in-process measurement technology, autocorrelated data are increasingly common in industrial SPC applications. A number of high performance control charting techniques that take into account the specific characteristics of the autocorrelation through time series modeling have been proposed over the past decade. We present a survey of such methods and analyze and compare their performances for a range of typical autocorrelated process models. One practical concern with these methods is that their performances are often strongly affected by errors in the time series models used to represent the autocorrelation. We also provide some analytical results comparing the robustness of the various methods with respect to time series modeling errors.

• Communications for Statistical Applications and Methods
• /
• 제10권3호
• /
• pp.931-954
• /
• 2003

It is widely recognized that economic time series involved not only the linearities but also the non-linearities. In this paper, when the economic time series data have the nonlinear characteristics we propose the forecasts method using combinations of both forecasts from linear and nonlinear models. In empirical study, we compare the forecasting performance of 4 exchange rates models(AR, GARCH, AR+GARCH, Bilinear model) and combination of these forecasts for dairly Won/Dollar exchange rates returns. The combination method is selected by the estimated individual forecast errors using Monte Carlo simulations. And this study shows that the combined forecasts using unrestricted least squares method is performed substantially better than any other combined forecasts or individual forecasts.

• 전기학회논문지
• /
• 제65권6호
• /
• pp.918-922
• /
• 2016

As the number of wind farms operating has increased, the interest of the central unit commitment and dispatch for wind power has increased as well. Wind power forecast is necessary for effective power system management and operation with high wind power penetrations. This paper presents the centralized wind power forecasting method, which is a forecast to combine all wind farms in the area into one, using time series models. Also, this paper proposes a prediction model modified with wind forecast error compensation. To demonstrate the improvement of wind power forecasting accuracy, the proposed method is compared with persistence model and new reference model which are commonly used as reference in wind power forecasting using Jeju Island data. The results of case studies are presented to show the effectiveness of the proposed wind power forecasting method.

• 응용통계연구
• /
• 제16권2호
• /
• pp.365-376
• /
• 2003

본 연구에서는 비정규 시계열 자료에 관한 다양한 회귀모형을 고찰하고, 이들 모형의 선택 기준에 관하여 연구해 보았다. 모형 선택의 기준으로는 AIC (Akaike information criterion), BIC (Baysian information criterion) 그리고 우도비 검정을 확장 적용하였다. 또한, 실제의 Polio 자료분석을 통해 이를 적용해보았다.

• Journal of the Korean Data and Information Science Society
• /
• 제17권1호
• /
• pp.63-76
• /
• 2006

Marine production forecasting in fisheries is a crucial factor for managing and maintaining fishery resources. Thus this paper aims to generate a forecasting model of total marine production. The most generally method of time series model is to generate the most optimal single forecasting model. But the method could induce a different forecasting results when it does not properly infer a model To overcome the defect, I am trying to propose a single forecasting through multiple time series model. In other word, by comparing and integrating the output resulted from ARIMA and VAR model (which are typical method in a forecasting methodology), I tried to draw a forecasting. It is expected to produce more stable and delicate forecasting prospect than a single model. Through this, I generated 3 models on a yearly and monthly data basis and then here I present a forecasting from 2006 to 2010 through comparing and integrating 3 models. In conclusion, marine production is expected to show a decreasing tendency for the coming years.

• 한국환경과학회지
• /
• 제24권8호
• /
• pp.1023-1036
• /
• 2015

A reliable streamflow forecasting is essential for flood disaster prevention, reservoir operation, water supply and water resources management. This study proposes a hybrid model for river stage forecasting and investigates its accuracy. The proposed model is the wavelet packet-based artificial neural network(WPANN). Wavelet packet transform(WPT) module in WPANN model is employed to decompose an input time series into approximation and detail components. The decomposed time series are then used as inputs of artificial neural network(ANN) module in WPANN model. Based on model performance indexes, WPANN models are found to produce better efficiency than ANN model. WPANN-sym10 model yields the best performance among all other models. It is found that WPT improves the accuracy of ANN model. The results obtained from this study indicate that the conjunction of WPT and ANN can improve the efficiency of ANN model and can be a potential tool for forecasting river stage more accurately.

• 전력전자학회논문지
• /
• 제27권4호
• /
• pp.316-324
• /
• 2022

This paper uses seasonal auto-regressive integrated moving average (S-ARIMA), which is efficient in seasonality between time-series models, to predict the degradation tendency for lithium-ion batteries and study a method for improving the predictive performance. The proposed method analyzes the degradation tendency and extracted factors through an electrical characteristic experiment of lithium-ion batteries, and verifies whether time-series data are suitable for the S-ARIMA model through several statistical analysis techniques. Finally, prediction of battery aging is performed through S-ARIMA, and performance of the model is verified through error comparison of predictions through mean absolute error.

• 한국시뮬레이션학회논문지
• /
• 제5권1호
• /
• pp.1-12
• /
• 1996

We propose a new approach to classifying a time series data into one of the autoregressive moving-average (ARMA) models. It is bases on two pattern recognition concepts for solving time series identification. The one is an extended sample autocorrelation function (ESACF). The other is a neural network-driven decision tree classifier(NNDTC) in which two pattern recognition techniques are tightly coupled : neural network and decision tree classfier. NNDTc consists of a set of nodes at which neural network-driven decision making is made whether the connecting subtrees should be pruned or not. Therefore, time series identification problem can be stated as solving a set of local decisions at nodes. The decision values of the nodes are provided by neural network functions attached to the corresponding nodes. Experimental results with a set of test data and real time series data show that the proposed approach can efficiently identify the time seires patterns with high precision compared to the previous approaches.

• 한국데이터정보과학회:학술대회논문집
• /
• 한국데이터정보과학회 2004년도 춘계학술대회
• /
• pp.91-101
• /
• 2004

The turbulent flow is of fundamental interest because the conservation equations for thermodynamics, mass and momentum are linked together. This turbulent flow consists of some coherent time- and space-organized vortical structures. Research has already shown that some dynamic systems and experimental models still cannot provide a good nonlinear analysis of turbulent time series. In the real turbulent flow, very complicated nonlinear behaviors, which are affected by many vague factors are present. In this paper, a kernel-based machine for fuzzy nonlinear regression analysis is proposed to predict the nonlinear time series of turbulent flows. In order to show the practicality and usefulness of this model, we present an example of predicting the near-wall turbulence time series as a verifiable model and compare with fuzzy piecewise regression. The results of practical applications show that the proposed method is appropriate and appears to be useful in nonlinear analysis and in fuzzy environments to predict the turbulence time series.

• 지능정보연구
• /
• 제24권4호
• /
• pp.1-32
• /
• 2018

본 연구는 경제적으로 국내에 큰 영향을 주었던 글로벌 금융위기를 기반으로 총 10년의 연간 기업데이터를 이용한다. 먼저 시대 변화 흐름에 일관성있는 부도 모형을 구축하는 것을 목표로 금융위기 이전(2000~2006년)의 데이터를 학습한다. 이후 매개 변수 튜닝을 통해 금융위기 기간이 포함(2007~2008년)된 유효성 검증 데이터가 학습데이터의 결과와 비슷한 양상을 보이고, 우수한 예측력을 가지도록 조정한다. 이후 학습 및 유효성 검증 데이터를 통합(2000~2008년)하여 유효성 검증 때와 같은 매개변수를 적용하여 모형을 재구축하고, 결과적으로 최종 학습된 모형을 기반으로 시험 데이터(2009년) 결과를 바탕으로 딥러닝 시계열 알고리즘 기반의 기업부도예측 모형이 유용함을 검증한다. 부도에 대한 정의는 Lee(2015) 연구와 동일하게 기업의 상장폐지 사유들 중 실적이 부진했던 경우를 부도로 선정한다. 독립변수의 경우, 기존 선행연구에서 이용되었던 재무비율 변수를 비롯한 기타 재무정보를 포함한다. 이후 최적의 변수군을 선별하는 방식으로 다변량 판별분석, 로짓 모형, 그리고 Lasso 회귀분석 모형을 이용한다. 기업부도예측 모형 방법론으로는 Altman(1968)이 제시했던 다중판별분석 모형, Ohlson(1980)이 제시한 로짓모형, 그리고 비시계열 기계학습 기반 부도예측모형과 딥러닝 시계열 알고리즘을 이용한다. 기업 데이터의 경우, '비선형적인 변수들', 변수들의 '다중 공선성 문제', 그리고 '데이터 수 부족'이란 한계점이 존재한다. 이에 로짓 모형은 '비선형성'을, Lasso 회귀분석 모형은 '다중 공선성 문제'를 해결하고, 가변적인 데이터 생성 방식을 이용하는 딥러닝 시계열 알고리즘을 접목함으로서 데이터 수가 부족한 점을 보완하여 연구를 진행한다. 현 정부를 비롯한 해외 정부에서는 4차 산업혁명을 통해 국가 및 사회의 시스템, 일상생활 전반을 아우르기 위해 힘쓰고 있다. 즉, 현재는 다양한 산업에 이르러 빅데이터를 이용한 딥러닝 연구가 활발히 진행되고 있지만, 금융 산업을 위한 연구분야는 아직도 미비하다. 따라서 이 연구는 기업 부도에 관하여 딥러닝 시계열 알고리즘 분석을 진행한 초기 논문으로서, 금융 데이터와 딥러닝 시계열 알고리즘을 접목한 연구를 시작하는 비 전공자에게 비교분석 자료로 쓰이기를 바란다.