• 응용통계연구
• /
• 제35권3호
• /
• pp.395-406
• /
• 2022

본 논문에서는 장기 종속 다변량 시계열 자료에 대한 이상점 탐지 기법을 연구한다. 기존 다변량 시계열 이상점 탐지 방법은 단기 종속 시계열 모형인 VARMA에 기반한 방법으로, 장기억성을 띈 다변량 시계열 자료에는 적합하지 않다. 자기회귀 모형을 통해서 장기 종속성, 즉 장기억성을 고려하기 위해서는 높은 차수의 모형이 필요하고, 이는 곧 추정의 불안성으로 이어지기에 장기억성을 효율적으로 다룰 수 없기 때문이다. 따라서, 본 논문은 이러한 문제를 보완하고자 VHAR 구조에 기반한 이상점 탐지 방법을 제시하고자 한다. 또한 더욱 정확한 추론을 위해서 로버스트한 방법을 이용하여 VHAR 계수를 추정하였고 이를 활용하여 이상점을 탐지하였다. 모의실험 결과 우리가 제안한 방법론이 기존 VARMA에 기반한 방법론보다 이상점 탐지에 더 효과적임을 살펴볼 수 있었다. 주가지수에 대한 실증자료 분석에서도 기존의 방법론은 탐지하지 못하는 추가 이상점을 찾음을 확인할 수 있었다.

• 자원ㆍ환경경제연구
• /
• 제9권3호
• /
• pp.515-544
• /
• 2000

This paper shows that various fractionally integrated univariate and multivariate are remarkably successful in representing annual temperature series and also very long series of tree ring widths, which are often used as a proxy for temperature. The analysis also suggests that human recorded temperature series are not inconsistent with being generated by a stationary, long memory process. From the empirical results, we should be noted that the statistically significant positive trend coefficients may well be due to small sample sizes. These results cast some doubt on the basic assumption that global warming is definitely occurring.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2021년도 학술발표회
• /
• pp.373-373
• /
• 2021

Rainfall forecasting is an important issue that is applied in many areas, such as agriculture, flood warning, and water resources management. In this context, this study proposed a statistical and machine learning-based forecasting model for monthly rainfall. The Bayesian Gaussian process was chosen to optimize the hyperparameters of the Stacked Long Short-term memory (SLSTM) model. The proposed SLSTM model was applied for predicting monthly precipitation of Seoul station, South Korea. Data were retrieved from the Korea Meteorological Administration (KMA) in the period between 1960 and 2019. Four schemes were examined in this study: (i) prediction with only rainfall; (ii) with deseasonalized rainfall; (iii) with rainfall and minimum temperature; (iv) with deseasonalized rainfall and minimum temperature. The error of predicted rainfall based on the root mean squared error (RMSE), 16-17 mm, is relatively small compared with the average monthly rainfall at Seoul station is 117mm. The results showed scheme (iv) gives the best prediction result. Therefore, this approach is more straightforward than the hydrological and hydraulic models, which request much more input data. The result indicated that a deep learning network could be applied successfully in the hydrology field. Overall, the proposed method is promising, given a good solution for rainfall prediction.