전 세계적인 물 부족 심화로 상수관망의 효율적 운영 및 유지보수(O&M) 중요성이 커지고 있다. 특히 정확한 수압 예측은 잠재적 문제의 사전 감지와 대응에 필수적이다. 이에 본 연구는 전처리된 데이터를 활용하여 현장 적용성이 높은 수압 예측 모델을 개발하는 것을 목표로 하였다. 이를 위해 8개 블록시스템(DMA)의 10분 단위 시계열 데이터와 4종류의 딥러닝 모델(LSTM, GRU, CNN-LSTM, CNN-GRU)을 활용하였으며, optuna를 통해 하이퍼파라미터를 최적화하고 배치 정규화 등을 적용해 학습 안정성을 확보하였다. 평가 결과, CNN-GRU 모델이 가장 우수한 성능을 나타냈다. 해당 모델을 기반으로 입력 조건에 따른 성능을 비교한 결과, 단변수 대비 다변수 입력 조건에서 예측 정확도가 향상됨을 확인하였다. 또한, 10분 선행 시점에서 최고 신뢰도(R2 0.9678, RMSE 0.0375)를 기록했으며, 지속성 모델의 성능이 점진적으로 하락하여 상대적인 저점을 형성하는 7시간 및 17시간 선행 시점에서 CNN-GRU 모델은 지속성 모델 대비 RMSE 기준 각각 48.0% 및 42.1%의 오차 개선을 달성하였다. 결론적으로, 본 연구에서 제안하는 전처리 및 하이퍼파라미터 통합 최적화 프로세스는 DMA별로 상이한 운영 환경에서도 안정적인 예측 성능을 확보할 수 있음을 입증하였다. 이는 현장 엔지니어의 데이터 분석 및 의사결정을 지원함으로써, 상수관망의 안정적인 운영과 유지보수 효율성 향상에 기여할 수 있을 것으로 기대된다.

최근 국민 삶의질 향상, 인구구조 변화 등에 따라 관광수요가 증가하고 관광활동 또한 다양화되고 있다. 특히 국가어항은 단순 한 어업 활동의 거점을 넘어 지역 경제, 관광, 문화까지 아우르는 복합 기능 공간으로 활용되고 있다. 본 연구는 부산 기장군에 위치한 대 변항을 대상으로 관광수요를 예측함으로써 정책적 활용이 가능한 기초자료를 제공하고자 한다. 2015년부터 2024년까지의 월별 위치 기반 방문객수를 입력 데이터로 설정하여 시계열 예측을 수행하였다. 연구방법론으로 전통적인 통계방법인 SARIMA를 기준으로 예측 정확도 를 향상시키기 위해 Hybrid model을 활용하였다. 특히 기존의 선형적 방법과 비선형 방법을 결합한 Hybrid model을 제안하고자 한다. 시계 열 구조적 분해방법인 STL 기법과 비구조적 잔차 제거 방법인 DAE 분석을 수행하였다. 즉 Trend와 Seasonal을 분해하는 STL과 머신 러닝 기반의 DAE를 활용하여 분해하고, 설명되지 않은 잔차를 대상으로 딥 러닝을 통해 예측함으로써 선형과 비선형 방법을 결합하여 예측 정확도를 제고하고자 한다. 분석결과, 연구진이 제안한 모델로 STL와 DAE를 활용하여 2중 분해하고 LSTM과 Attention을 결합한 Hybrid deep learning model이 가장 예측 정확도가 높았다. 향후 관광수요 예측에서는 분해와 잔차 기반의 노이즈 제거 과정을 거친 후, 딥 러닝 기 법을 결합하는 것이 정확성 측면에서 효과적인 방법임을 확인하였다. 본 연구는 연안 국가어항의 미래 관광수요를 전망함으로써, 관광추 세를 반영한 지속가능한 수요 대응 전략 수립 및 정책 의사결정에 기여할 수 있을 것으로 기대된다.

고자리꽃파리는 양파 및 마늘 등 Allium 속에 속하는 농작물의 중요한 해충으로 전 세계적으로 온대 지역에서 경제적 해충으로 취급하고 있다. 본 연구는 고자리꽃파리의 발생 기준점을 정하여 연간발생양상을 해석하고, 초기방제 시기를 설정할 수 있도록 월동 번데기 우화모형을 개발하고자 수행하였다. 고자리꽃파리 월동 번데기의 온도발육 모형으로 선형 및 비선형 모형을 추정하고, 발육기간 분포모형과 결합하여 월동번데기의 성충으로 우화시기 예측모형을 수립하였다. 비선형 모델의 경우 3-매개변수 Lactin 수식과 저온에서 온도와 발육률 간의 선형성을 높이기 위해 마지막 매개변수 (λ)를 선형모형의 절편으로 대체한 4-매개변수 수식을 사용하였다. 일일 평균기온을 이용하여 50% 성충 우화일을 예측한 결과, 선형모형 기반의 적 산온도 모형(DD, degree-day) 및 선형 또는 비선형 모형을 적용하여 발육률을 누적하는 발육률 적산 모형(RS, rate summation) 모두 실측값과 큰 차이를 보였다. 반면 시간별 온도를 입력변수를 사용한 경우, 3-매개변수 모델을 제외한 사인곡선법 기반의 DD 모형, 선형 RS 모형, 4-매개변수 비선 형 RS 모형의 평균편차는, 실제 관측치와 3일 이상 차이가 나지 않았다. 최종적으로 시간별 온도자료를 이용하고, 발육모형으로 선형과 4-매개변수 비 선형 모형을 적용하는 RS 모형을 활용 가능한 모형으로 선정하였다. 선형 RS 모형은 두 번의 포장적합(1984, 1987)에서 실제 관측값과 편차가 3일 이내로 차이가 없었다. 비선형 RS 모형은 1984년 적합에서 0.8일의 편차로 정확했지만 1987년 적합에서는 6.5일의 평균편차를 보였다.

복숭아유리나방(Synanthedon bicingulata)은 유충이 나무 줄기 속으로 파고 들어가 형성층을 섭식하는 생태적 특징으로 인해 방제가 어려운 해충이다. 현재에는 성충의 발생시기를 지속적인 모니터링을 통해 화학 방제에 의존하고 있다. 따라서, 성충 발생 시기를 예측할 수 있다면 표본조사와 방제 효율을 극대화할 것으로 기대된다. 본 연구에서는 국내에서 발표된 발생 소장 연구의 발생 데이터와 해당 지역의 기온 데이터를 활용, Weibull function 을 이용하여 복숭아유리나방의 성충 발생 최성기를 예측하는 모델을 개발하였다. 또한 개발된 모델과 SSP 미래 기후변화 시나리오를 이용해 미래 기후변화 상황에서 복숭아유리나방의 전국적 발생 양상이 어떻게 변화할지 예측해보았다. 복숭아유리나방의 성충 발생은 온일도일에 따라 예측이 가능하였고 연 중 2회의 성충 최성기가 발생하는 것으로 예측되었다. 이번 연구에서 개발된 모델은 첫 번째와 두 번째 성충 최성기(50% 발생시기)를 국내 전역에서 평균 6.3일, 4.0일 이내로 예측해 예측 정확도가 매우 높았다. 이번 연구 결과는 난방제 해충인 복숭 아유리나방의 방제 효율을 급격히 높혀줄 뿐만아니라, 기후변화에 따른 복숭아유리나방의 발생 변화 예측에도 기여할 수 있을 것으로 기대된다.

This study endeavors to enrich investment prospects in cryptocurrency by establishing a rationale for investment decisions. The primary objective involves evaluating the predictability of four prominent cryptocurrencies – Bitcoin, Ethereum, Litecoin, and EOS – and scrutinizing the efficacy of trading strategies developed based on the prediction model. To identify the most effective prediction model for each cryptocurrency annually, we employed three methodologies – AutoRegressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM), and Prophet – representing traditional statistics and artificial intelligence. These methods were applied across diverse periods and time intervals. The result suggested that Prophet trained on the previous 28 days' price history at 15-minute intervals generally yielded the highest performance. The results were validated through a random selection of 100 days (20 target dates per year) spanning from January 1st, 2018, to December 31st, 2022. The trading strategies were formulated based on the optimal-performing prediction model, grounded in the simple principle of assigning greater weight to more predictable assets. When the forecasting model indicates an upward trend, it is recommended to acquire the cryptocurrency with the investment amount determined by its performance. Experimental results consistently demonstrated that the proposed trading strategy yields higher returns compared to an equal portfolio employing a buy-and-hold strategy. The cryptocurrency trading model introduced in this paper carries two significant implications. Firstly, it facilitates the evolution of cryptocurrencies from speculative assets to investment instruments. Secondly, it plays a crucial role in advancing deep learning- based investment strategies by providing sound evidence for portfolio allocation. This addresses the black box issue, a notable weakness in deep learning, offering increased transparency to the model.

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.

Seaports play a huge role in Vietnam's economy, being a border gate for economic and cultural exchanges with outsiders, especially the role of goods circulation. Container cargoes are one of the types of goods with large and increasing volume through Vietnam seaports. However, the heterogeneity between the seaport and the connected infrastructure greatly affects the capacity and efficiency of port investment. This is also one of the main causes leading to a shortage of goods, excess ports in some port areas. The root cause is that the planning has not kept up with the growth of the amount of goods arriving at the port, because the issue of forecasting the volume of goods through the port is not accurate. Therefore, it is necessary to develop models of forecasting container cargo through the ports with general, scientific, and high accuracy to serve the strategy, planning and development of seaport system; the work of planning and investment in the development of seaports, shipping fleets and other auxiliary transport infrastructure works. The purpose of this study is to build suitable forecasting models with high accuracy and reliability on the total volume of container cargo throughput of the Vietnamese seaport system. Based on the methods of a statistical survey, synthesis, regression analysis, and correlation to evaluate the influence of factors on container cargo volume through Vietnam's seaports in the period of 2004-2019. By incorporating more economic factors into the regression model, the paper focuses on forecasting container cargo through the Vietnamese seaport systems, going into cargo-based forecasting in tons and TEUs. The results of this study contribute to complete the rationale for forecasting, especially forecasts related to the shipping industry and the forecast for container cargo throughput of the seaport. Finally, selecting models for forecasting container cargo volume throughput of seaports by Vietnamese conditions.

Recently, the continuing operation of nuclear power plants has become a major controversial issue in Korea. Whether to continue to operate nuclear power plants is a matter to be determined considering many factors including social and political factors as well as economic factors. But in this paper we concentrate only on the economic factors to make an optimum decision on operating nuclear power plants. Decisions should be based on forecasts of plant accident risks and large and small accident data from power plants. We outline the structure of a decision model that incorporate accident risks. We formulate to decide whether to shutdown permanently, shutdown temporarily for maintenance, or to operate one period of time and then periodically repeat the analysis and decision process with additional information about new costs and risks. The forecasting model to predict nuclear power plant accidents is incorporated for an improved decision making. First, we build a one-period decision model and extend this theory to a multi-period model. In this paper we utilize influence diagrams as well as decision trees for modeling. And bayesian statistical approach is utilized. Many of the parameter values in this model may be set fairly subjective by decision makers. Once the parameter values have been determined, the model will be able to present the optimal decision according to that value.

There are various issues affecting the financial revenues of expressways, such as new transportation systems, the opening of new roads, and free toll charges. As a result, expressway toll revenues for 2017 increased only 0.3% from the previous year. If this trend continues, the steady increase in expressway revenue may have occurred, therefore it is necessary to improve the model of annual trips and revenues considering various external variables that are occurring recently. In this study, we constructed annual trips forecasting model that can analyze more precisely the changes of road network by using new independent variables (which are not considered in existing models) such as weighted length considering regional traffic volume level and private road ratio. Also we performed a basic statistical analysis on oil prices and reflected the model as a dummy variable to improve the explanatory power of the model. And we established an analysis process to estimate the toll revenue. The results of this study can be used as the basic data for expressway financial model.

비래해충인 혹명나방(Cnaphalocrocis medinalis)과 멸강나방(Mythimna separata)은 아시아의 주요 벼 재배국가에 광범위하게 분포하고 있는 벼의 주요 해충이다. 국내에서는 벼멸구, 흰등멸구와 함께 중국에서 비래하여 나타나는 것으로 알려져 있다. 혹명나방과 멸강나방의 발생지역과 통계적으로 유의미한 상관관계에 있는 환경변수를 확인하고, 국내에서의 지속적인 발생 가능성을 알아보기 위해 Maxent (Maximum Entropy Model) 3.3.2를 사용하였다.

This study builds counties-specific panel data and establish a stochastic rice yield forecasting model by using a fixed effect panel model based on results calculating the coefficients for the meteorological factors, and by using a variety of weather scenarios. Rice yield prediction model developed estimating equations were set to rice yield as the dependent variable, and the average temperature, accumulated temperature, daily temperature range, sunshine hours as explanatory variables, by using panel data by counties in recent 10 years. Estimation results using a fixed-effects model was able to verify that an average temperature affects to yield as quadratic form, there appeared to be significantly affected by accumulated temperature in Heading period, an average temperature in Ripening period. a rice yield prediction model is meaningful in that we can see the forecasting results in the previous. not waiting the actual survey results provided by the National Statistical Office. because this forecasting estimates is sufficient rationale material by government supply & demand measures. Finally, the study leave to future challenges with respect to establishing a prediction model developed as combined with land productivity and environmental engineering factors.

Fashion is primarily based on adoption of trends by consumers in textiles, clothing, footwear, jewelry and art, inter alia. As fashion is based on human preferences, it is characterized by dynamic changes throughout seasons and years, short life cycles, low predictability and high volatility of demand and impulse purchases. In the dynamic environment of apparel markets, fashion firms aim at successfully forecasting both the desirability of new collections and the volumes of each item produced and released to the market under terms of substantial levels of uncertainty. When demand for an item exceeds its supply, the firm is likely to lose additional profits that could have been collected had a sufficient volume of this item been present in the market. Alternatively, if the supply of an item surpassed its demand, it would remain unsold, thereby generating loss equal to its marginal production and distribution costs. The paper proposes a forecasting model that enhances the accuracy of fashion trend forecasting in the context of multiple variants of colour clothing. The model aims at maximizing profits of the firms, while minimizing the forecasting error and reducing the costs that result from excess capacity of production or, alternatively, from loss of potential revenues due to low demand.

There have been various studies on measurements of flood risk and forecasting models. For river and dam region, PDF and FVI has been proposed for measurement of flood risk and regression models have been applied for forecasting model. For Bo region unlikely river or dam region, flood risk would unexpectedly increase due to outgoing water to keep water amount under the designated risk level even the drain system could hardly manage the water amount. GFI and general linear model was proposed for flood risk measurement and forecasting model. In this paper, FVI with the consideration of duration on GFI was proposed for flood risk measurement at Bo region. General linear model was applied to the empirical data from Bo region of Nadong river to derive the forecasting model of FVI at three different values of Base High Level, 2m, 2.5m and 3m. The significant predictor variables on the target variable, FVI were as follows: ground water level based on sea level with negative effect, difference between ground altitude of ground water and river level with negative effect, and difference between ground water level and river level after Bo water being filled with positive sign for quantitative variables. And for qualitative variable, effective soil depth and ground soil type were significant for FVI.

Spare part management is very important to products that have large number of parts and long lifecycle such as automobile and aircraft. Supply chain must support immediate procurement for repair. However, it is not easy to handle spare parts efficiently due to huge stock keeping units. Qualified forecasting is the basis for the supply chain to achieve the goal. In this paper, we propose an agent based modeling approach that can deal with various factors simultaneously without mathematical modeling. Simulation results show that the proposed method is reasonable to describe demand generation process, and consequently, to forecast demand of spare parts in long-term perspective.