• Communications for Statistical Applications and Methods
• /
• v.18 no.1
• /
• pp.13-21
• /
• 2011

This paper studies the analysis of multivariate nonstationary time series with seasonality. Three types of multivariate time series models are considered: seasonal cointegration model, nonseasonal cointegration model with seasonal dummies, and vector autoregressive model in seasonal differences that are compared for forecasting performances using Korean macro-economic time series data. The cointegration models produce smaller forecast errors in short horizons; however, when longer forecasting periods are considered the vector autoregressive model appears preferable.

• Journal of Korean Society of Transportation
• /
• v.25 no.1 s.94
• /
• pp.121-132
• /
• 2007

Up to now Permanent traffic volumes have been counted by Automatic Vehicle Classification (AVC) on National Highways. When counted data have missing items or errors, the data must be revised to stay statistically reliable This study was carried out to estimate correct data based on outoregression and seasonal AutoRegressive Integrated Moving Average (ARIMA). As a result of verification through seasonal ARIMA, the longer the missed period is, the greater the error. Autoregression results in better verification results than seasonal ARIMA. Traffic data is affected by the present state mote than past patterns. However. autoregression can be applied only to the cases where data include similar neighborhood patterns and even in this case. the data cannot be corrected when data are missing due to low qualify or errors Therefore, these data shoo)d be corrected using past patterns and seasonal ARIMA when the missing data occurs in short periods.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2006.05a
• /
• pp.23-26
• /
• 2006

기존의 데이터 마이닝 예측 기법 중 회귀분석은 학습 단계에서 생성된 모델을 변경 없이 새로운 데이터에 적용하였다. 그러나 시계열 데이터에 모델 변경 없이 동일하게 적용하면 시간이 지남에 따라 정확도가 낮아지는 단점이 있다. 따라서 이 논문에서는 시간에 따라 변화하는 시계열데이터의 특성을 고려하여 점진적으로 회귀 모델을 갱신하는 기법을 제안한다. 이 기법은 입력되는 모든 데이터를 회귀 모델에 적용하여 점진적으로 모델을 갱신한다. 제안된 기법의 타당성은 RME(Relative Mean Error)와 RMSE(Root Mean Square Error)를 이용하여 측정하였다. 정확도 측정 실험 결과 제안 기법인 IMQR(Incremental Multiple Quadratic Regression) 기법이 MLR(Multiple Linear Regression), MQR(Multiple Quadratic Regression), SVR(Support Vector Regression) 기법에 비해 RME 가 평균 2%, RMSE 가 평균 0.02 정도 우수한 결과를 얻었다.

• The Korean Journal of Applied Statistics
• /
• v.37 no.3
• /
• pp.297-309
• /
• 2024

This paper evaluates the performance of the neural network autoregressive model combined with an exponential smoothing model, called the NNARX+ETS model. The combined model utilizes the components of ETS as exogenous variables for NNARX, to forecast time series data using artificial neural networks. The main idea is to enhance the performance of NNAR using only lags of the original time series data, by combining traditional time series analysis methods with the neural networks through NNARX. We employ two real data for performance evaluation and compare the NNARX+ETS with NNAR and traditional time series analysis methods such as ETS and ARIMA (autoregressive integrated moving average) models.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.99-99
• /
• 2021

본 연구에서는 기상 자료와 적설 특성 자료의 관계를 도출하고, 이와 추계 일기 생성 모형을 활용하여 합성 적설심 시계열을 모의하는 방법에 대하여 제안한다. 추계 일기 생성 모형에서는 적설량을 직접 모의하지 않기 때문에 강수량을 적설량으로 변환해야한다. 이를 위해 도입한 관계식은 다음과 같다. 첫째로 기상청 적설 예보의 적설 유무 판단 기준을 이용하였다. 이 기준에서는 상대습도와 지상기온에 따라 강수의 형태를 비, 눈, 진눈깨비로 구분한다. 둘째로 강수가 적설로 판단되었을 때 강수량을 신적설심으로 환산하는 수상당량비를 지상기온과 회귀 분석하였다. 선행 연구에 따라 3시간 1 mm 이상 5 mm 이하 강수와 3시간 5 mm 이상 강수 사상에 대하여 나누어 sigmoid형 곡선을 이용하여 회귀 분석하였다. 마지막으로 융설에 의한 적설심 감소량을 지상기온과 복사량의 함수로 표현하였으며, 각 변수의 계수는 입자 군집 최적화 방법을 통하여 보정하였다. 추계 일기 생성 모형으로는 AWE-GEN 모형을 활용하였으며, 시험 자료로 강릉(105) 종관기상관측소의 24년 기간(1982-2005) 자료를 활용하여 합성 적설심 시계열을 생성하였다. 합성 적설심 시계열 모의 과정은 다음과 같다. (1) 추계 일기 생성 모형으로 합성 일기 자료 생성, (2) 강수 발생 시 적설 유무 판단, (3) 적설로 판단 시 수상당량비를 계산하여 신적설심 추정, (4) 기존 적설심에 신적설심을 더하고, 적설심 감소량만큼 감소. 위와 같은 과정으로 200년 길이 합성 적설심 시계열을 모의한 결과 극한 사상을 과소 추정하는 경향이 나타나 추가적인 개선이 필요한 것으로 판단된다.

• Information Systems Review
• /
• v.25 no.3
• /
• pp.163-177
• /
• 2023

This research aims to identify key initiatives and a policy approach to support the industrialization of the sector. The research collected a total of 91,873 news data points relating to healthcare between 2013 to 2022. A total of 20 topics were derived through topic modeling analysis, and as a result of time series regression analysis, 4 hot topics (Healthcare, Biopharmaceuticals, Corporate outlook·Sales, Government·Policy), 3 cold topics (Smart devices, Stocks·Investment, Urban development·Construction) derived a significant topic. The research findings will serve as an important data source for government institutions that are engaged in the formulation and implementation of Korea's policies.

• The Korean Journal of Applied Statistics
• /
• v.16 no.2
• /
• pp.365-376
• /
• 2003

In this paper we reviewed a variety of non-Gaussian time series models, and studied the model selection criteria such as AIC and BIC to select proper models. We also considered the likelihood ratio test and applied it to analysis of Polio data set.

• Journal of the Korean Data and Information Science Society
• /
• v.24 no.6
• /
• pp.1103-1112
• /
• 2013

Exploratory data analysis focuses mostly on data exploration instead of model fitting. We can use the recurrence plot as a graphical exploratory data analysis tool. With the recurrence plot, we can obtain the structural pattern of the time series and recognize the structural change points in time series at a glance.

• Journal of Digital Convergence
• /
• v.10 no.9
• /
• pp.357-362
• /
• 2012

We have compared and predicted for non-linear time series data which are real data having different variences using GRCA(1) model and neural network method. In particular, using Korea Composite Stock Price Index rate, mean square errors of prediction are obtained in genaralized random coefficient autoregressive model and neural network method. Neural network method prove to be better in short-term forecasting, however GRCA(1) model perform well in long-term forecasting.

• Journal of the Korean Data and Information Science Society
• /
• v.20 no.5
• /
• pp.857-867
• /
• 2009

The purpose of this paper is to suggest the methods of forecasting for the number of the elementary, middle and high-school student based on the proportion of promotion until 2026 year. The suggested methods are the proportion of promotion, mov baseverage, Holt-W bters model, SARIMA, regression fit. As the result, the abilities of forecasting by the method of moving average are better than those of other methods.