• 지능정보연구
• /
• 제19권1호
• /
• pp.1-17
• /
• 2013

본 연구에서는 시계열 예측을 위해 선형 모형과 비선형 모형의 하이브리드 모형 및 순수 모형의 성과를 비교 평가하였다. 이를 위해 5가지 서로 다른 패턴을 가지는 데이터를 생성하여 시뮬레이션을 진행하였다. 본 연구에서 고려한 선형 모형은 AR(autoregressive model)과 SARIMA(seasonal autoregressive integrated moving average model)이고 비선형 모형은 인공신경망(artificial neural networks model)과 GAM(generalized additive model)이다. 특히, GAM은 여러 장점에도 불구하고 시계열 예측을 위한 비선형 모형으로 기존 연구들에서는 거의 쓰이지 않았던 모형이다. 시뮬레이션 결과, seasonality를 가지는 시계열에 대해서는 AR 및 AR-AR 모형이, trend를 가지는 시계열에 대해서는 SARIMA 및 SARIMA와 다른 모형의 하이브리드 모형이 다른 모형에 비해 높은 성과를 보였다. 한편, 인공신경망과 GAM을 비교하면, 트렌드와 계절성이 더해진 시계열에 대해 SARIMA와 GAM의 하이브리드 모형이 거의 모든 노이즈(noise) 수준에 대해 높은 성과를 보인 반면, 노이즈 수준이 미미한 경우에 한해 SARIMA와 인공신경망의 하이브리드 모형이 높은 성과를 보였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권2호
• /
• pp.383-403
• /
• 2021

With the development of mobile edge computing (MEC), some late-model application technologies, such as self-driving, augmented reality (AR) and traffic perception, emerge as the times require. Nevertheless, the high-latency and low-reliability of the traditional cloud computing solutions are difficult to meet the requirement of growing smart cars (SCs) with computing-intensive applications. Hence, this paper studies an efficient offloading decision and resource allocation scheme in collaborative vehicular edge computing networks with multiple SCs and multiple MEC servers to reduce latency. To solve this problem with effect, we propose a context-aware offloading strategy based on differential evolution algorithm (DE) by considering vehicle mobility, roadside units (RSUs) coverage, vehicle priority. On this basis, an autoregressive integrated moving average (ARIMA) model is employed to predict idle computing resources according to the base station traffic in different periods. Simulation results demonstrate that the practical performance of the context-aware vehicular task offloading (CAVTO) optimization scheme could reduce the system delay significantly.

• 정보과학회 논문지
• /
• 제44권9호
• /
• pp.954-965
• /
• 2017

안정적인 전력 공급은 전력 인프라의 유지 보수 및 작동에 매우 중요하며, 이를 위해 정확한 전력 사용량 예측이 요구된다. 대학 캠퍼스는 전력 사용량이 많은 곳이며, 시간과 환경에 따른 전력 사용량 변화폭이 다양하다. 이러한 이유로, 전력계통의 효율적인 운영을 위해서는 전력 사용량을 정확하게 예측할 수 있는 모델이 요구된다. 기존의 시계열 예측 기법은 학습 시점과 예측 시점 간의 차이가 클수록 예측 구간이 넓어짐으로 예측 성능이 크게 떨어진다는 단점이 있다. 본 논문은 이를 보완하려는 방안으로, 먼저 의사결정나무를 이용해 날짜, 요일, 공휴일 여부, 학기 등을 고려하여 시계열 형태가 유사한 전력 데이터를 분류한다. 다음으로 분류된 데이터 셋에 각각의 자기회귀누적이동평균모형을 구성하여, 예측 시점에서 시계열 교차검증을 적용해 대학 캠퍼스의 일간 전력 사용량 예측 기법을 제안한다. 예측의 정확성을 평가하기 위해, 성능 평가 지표를 이용하여 제안한 기법의 타당성을 검증하였다.

• Journal of Preventive Medicine and Public Health
• /
• 제44권1호
• /
• pp.2-8
• /
• 2011

Objectives: Public release of and feedback (here after public release) on institutional (clinics and hospitals) cesarean section rates has had the effect of reducing cesarean section rates. However, compared to the isolated intervention, there was scant evidence of the effect of repeated public releases (RPR) on cesarean section rates. The objectives of this study were to evaluate the effect of RPR for reducing cesarean section rates. Methods: From January 2003 to July 2007, the nationwide monthly institutional cesarean section rates data (1 951 303 deliveries at 1194 institutions) were analyzed. We used autoregressive integrated moving average (ARIMA) time-series intervention models to assess the effect of the RPR on cesarean section rates and ordinal logistic regression model to determine the characteristics of the change in cesarean section rates. Results: Among four RPR, we found that only the first one (August 29, 2005) decreased the cesarean section rate (by 0.81 percent) and continued to have an impact period through the last observation in May 2007. Baseline cesarean section rates (OR, 4.7; 95% CI, 3.1 to 7.1) and annual number of deliveries (OR, 2.8; 95% CI, 1.6 to 4.7) of institutions in the upper third of each category at before first intervention had a significant contribution to the decrease of cesarean section rates. Conclusions: We could not found the evidence that RPR has had the significant effect of reducing cesarean section rates. Institutions with upper baseline cesarean section rates and annual number of deliveries were more responsive to RPR.

• 보건행정학회지
• /
• 제22권3호
• /
• pp.297-314
• /
• 2012

Purpose : The purpose of this study was to analyze the trend of the number of nursing staffs and skill mix and to assess the effectiveness of hospital nurse expansion policies in Korea. Methods : The trend of the number of nursing staffs and skill mix were analyzed using time series data, which composed of yearly series data from 1975 to 2009. The impact of hospital nurse expansion policies was estimated by autoregressive integrated moving average(ARIMA) intervention model. Results : The number of general hospital and hospital nurses per 100 beds was decreased in late 1980s and late 1990s due to rapid growth of beds. As a result of the number of nurse aids per 100 beds decreased, skill mix became high in general hospital but nurse ratio among hospital nursing staffs was about 50%. Expansion of new nurse and revised differentiated inpatient fee were only effective in expansion of hospital nursing staffs. But they had no effect in general hospitals. Conclusion : In Korea, a few policies related to expansion of hospital nurses have an effect on increasing the number of hospital nurse. Nevertheless, level of hospital nursing staffs is inferior to that of general hospital.

• 응용통계연구
• /
• 제32권1호
• /
• pp.69-81
• /
• 2019

본 논문에서는 딥러닝을 이용한 이변량 장기종속시계열(long-range dependent time series) 예측을 고려하였다. 시계열 데이터 예측에 적합한 LSTM(long short-term memory) 네트워크를 이용하여 이변량 장기종속시계열을 예측하고 이를 이변량 FARIMA(fractional ARIMA) 모형인 FIVARMA 모형과 VARFIMA 모형과의 예측 성능을 실증 자료 분석을 통해 비교하였다. 실증 자료로는 기능적 자기공명 영상(fMRI) 및 일일 실현 변동성(daily realized volatility) 자료를 이용하였으며 표본외 예측(out-of sample forecasting) 오차 비교를 통해 예측 성능을 측정하였다. 그 결과, FIVARMA 모형과 VARFIMA 모형의 예측값에는 미묘한 차이가 존재하며, LSTM 네트워크의 경우 초매개변수 선택으로 복잡해 보이지만 계산적으로 더 안정되면서 예측 성능도 모수적 장기종속시계열과 뒤지지 않은 좋은 예측 성능을 보였다.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2021년도 학술발표회
• /
• pp.326-326
• /
• 2021

장기간에 걸쳐 넓은 지역에 대해 발생하는 가뭄을 예측하기위해 많은 학자들의 기술적, 학술적 시도가 있어왔다. 본 연구에서는 복잡한 시계열을 가진 가뭄을 전망하는 방법 중 시나리오에 기반을 둔 가뭄전망 방법과 실시간으로 가뭄을 예측하는 비시나리오 기반의 방법 등을 이용하여 미래 가뭄전망을 실시했다. 시나리오에 기반을 둔 가뭄전망 방법으로는, 3개월 GCM(General Circulation Model) 예측 결과를 바탕으로 2009년도 PDSI(Palmer Drought Severity Index) 가뭄지수를 산정하여 가뭄심도에 대한 단기예측을 실시하였다. 또, 통계학적 방법과 물리적 모델(Physical model)에 기반을 둔 확정론적 수치해석 방법을 이용하여 비시나리오 기반 가뭄을 예측했다. 기존 가뭄을 통계학적 방법으로 예측하기 위해서 시도된 대표적인 방법으로 ARIMA(Autoregressive Integrated Moving Average) 모델의 예측에 대한 한계를 극복하기위해 서포트 벡터 회귀(support vector regression, SVR)와 웨이블릿(wavelet neural network) 신경망을 이용해 SPI를 측정하였다. 최적모델구조는 RMSE(root mean square error), MAE(mean absolute error) 및 R(correlation Coefficient)를 통해 선정하였고, 1-6개월의 선행예보 시간을 갖고 가뭄을 전망하였다. 그리고 SPI를 이용하여, 마코프 연쇄(Markov chain) 및 대수선형모델(log-linear model)을 적용하여 SPI기반 가뭄예측의 정확도를 검증하였으며, 터키의 아나톨리아(Anatolia) 지역을 대상으로 뉴로퍼지모델(Neuro-Fuzzy)을 적용하여 1964-2006년 기간의 월평균 강수량과 SPI를 바탕으로 가뭄을 예측하였다. 가뭄 빈도와 패턴이 불규칙적으로 변하며 지역별 강수량의 양극화가 심화됨에 따라 가뭄예측의 정확도를 높여야 하는 요구가 커지고 있다. 본 연구에서는 복잡하고 비선형성으로 이루어진 가뭄 패턴을 기상학적 가뭄의 정도를 나타내는 표준강수증발지수(SPEI, Standardized Precipitation Evapotranspiration Index)인 월SPEI와 일SPEI를 기계학습모델에 적용하여 예측개선 모형을 개발하고자 한다.

• 한국터널지하공간학회 논문집
• /
• 제21권2호
• /
• pp.227-242
• /
• 2019

토압식(Earth Pressure-Balanced, EPB) 쉴드 TBM 기계데이터 분석을 통해 토사터널의 특징이 반영된 막장 전방 예측 방법을 제안하였다. 기존에 암반과 토사가 혼합된 복합 지반의 예측에 적용하였던 시계열 분석 모델을 토사터널에 적용가능하도록 수정하였다. 또한 수정된 모델을 사용하여, 토사 종류에 따라 쏘일 컨디셔닝 재료를 선택하는 것이 타당한지 연구하였다. 이를 위해 Self-Organizing Map (SOM) 군집화(clustering) 분석을 수행하였다. 그 결과 무엇보다도 지반타입이 #200체 통과량 35% 기준으로 분류되어야 한다는 것을 확인하였다. 또한 TBM 기계데이터 분석을 통해 수정된 모델이 지반 타입을 예측하는데 사용될 수 있음을 확인하였다. 수정된 기준에 따라 지반 타입을 분류하고 시계열 분석을 수행하면, 10막장 전방 지반에 대해서 98%의 높은 예측 정확도를 보였으며, 이를 통해 수정된 방법의 우수성이 입증되었다. 특히 지반 타입 변화 구간에 대한 예측 정확도도 약 93%로, 10막장 전방에서 지반 타입 변화 여부를 미리 확인할 수 있게 되었다.

• 산경연구논집
• /
• 제11권8호
• /
• pp.31-38
• /
• 2020

Purpose: This study uses 'Autoregressive Integrated Moving Average Model' to predict the impact of a sharp drop in the base rate due to COVID-19 at the present time when government policies for stabilizing house prices are in progress. The purpose of this study is to predict implications for the direction of the government's house policy by predicting changes in house transaction prices and house rental prices after a sharp cut in the base rate. Research design, data, and methodology: The ARIMA intervention model can build a model without additional information with just one time series. Therefore, it is a time-series analysis method frequently used for short-term prediction. After the subprime mortgage, which had shocked since the global financial crisis in April 2007, the bank's interest rate in 2020 is set at a time point close to zero at 0.75%. After that, the model was estimated using the interest rate fluctuations for the Bank of Korea base interest rate, the house transaction price index, and the house rental price index as event variables. Results: In predicting the change in house transaction price due to interest rate intervention, the house transaction price index due to the fall in interest rates was predicted to change after 3 months. As a result, it was 102.47 in April 2020, 102.87 in May 2020, and 103.21 in June 2020. It was expected to rise in the short term. In forecasting the change in house rental price due to interest rate intervention, the house rental price index due to the drop in interest rate was predicted to change after 3 months. As a result, it was 97.76 in April 2020, 97.85 in May 2020, and 97.97 in June 2020. It was expected to rise in the short term. Conclusions: If low interest rates continue to stimulate the contracted economy caused by COVID-19, it seems that there is ample room for house transaction and rental prices to rise amid low growth. Therefore, In order to stabilize the house price due to the low interest rate situation, it is considered that additional measures are needed to suppress speculative demand.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권4호
• /
• pp.2060-2077
• /
• 2019

Recently, mobile healthcare services have attracted significant attention because of the emerging development and supply of diverse wearable devices. Smartwatches and health bands are the most common type of mobile-based wearable devices and their market size is increasing considerably. However, simple value comparisons based on accumulated data have revealed certain problems, such as the standardized nature of health management and the lack of personalized health management service models. The convergence of information technology (IT) and biotechnology (BT) has shifted the medical paradigm from continuous health management and disease prevention to the development of a system that can be used to provide ground-based medical services regardless of the user's location. Moreover, the IT-BT convergence has necessitated the development of lifestyle improvement models and services that utilize big data analysis and machine learning to provide mobile healthcare-based personal health management and disease prevention information. Users' health data, which are specific as they change over time, are collected by different means according to the users' lifestyle and surrounding circumstances. In this paper, we propose a prediction model of user physical activity that uses data characteristics-based long short-term memory (DC-LSTM) recurrent neural networks (RNNs). To provide personalized services, the characteristics and surrounding circumstances of data collectable from mobile host devices were considered in the selection of variables for the model. The data characteristics considered were ease of collection, which represents whether or not variables are collectable, and frequency of occurrence, which represents whether or not changes made to input values constitute significant variables in terms of activity. The variables selected for providing personalized services were activity, weather, temperature, mean daily temperature, humidity, UV, fine dust, asthma and lung disease probability index, skin disease probability index, cadence, travel distance, mean heart rate, and sleep hours. The selected variables were classified according to the data characteristics. To predict activity, an LSTM RNN was built that uses the classified variables as input data and learns the dynamic characteristics of time series data. LSTM RNNs resolve the vanishing gradient problem that occurs in existing RNNs. They are classified into three different types according to data characteristics and constructed through connections among the LSTMs. The constructed neural network learns training data and predicts user activity. To evaluate the proposed model, the root mean square error (RMSE) was used in the performance evaluation of the user physical activity prediction method for which an autoregressive integrated moving average (ARIMA) model, a convolutional neural network (CNN), and an RNN were used. The results show that the proposed DC-LSTM RNN method yields an excellent mean RMSE value of 0.616. The proposed method is used for predicting significant activity considering the surrounding circumstances and user status utilizing the existing standardized activity prediction services. It can also be used to predict user physical activity and provide personalized healthcare based on the data collectable from mobile host devices.