• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.340-340
• /
• 2019

지구온난화로 유발되는 기후변동성이 증가함에 따라서 정확한 수문인자의 예측은 전 세계적으로 주요 관심사항이 되고 있다. 최근에는 고성능 컴퓨터 자원의 증가로 수문기상학 연구에서 동일한 학습량에 비하여 정확도의 향상이 뚜렷한 기계학습 구조를 활용하여 위성영상 기반의 대기예측, 태풍위치 추적 및 강수량 예측 등의 연구가 활발하게 진행되고 있다. 본 연구에는 기계학습 중 시계열 분석에 널리 활용되고 있는 순환신경망(Recurrent Neural Network, RNN) 기법의 대표적인 LSTM(Long Short-Term Memory) 네트워크를 이용하여 수문인자를 예측하였다. LSTM 네트워크는 가중치 및 메모리 요소에 대한 추가정보를 셀 상태에 저장하고 시계열의 길이 조정하여 모형의 탄력적 활용이 가능하다. LSTM 네트워크를 이용한 다양한 수문인자 예측결과 RMSE의 개선을 확인하였다. 따라서 본 연구를 통하여 개발된 기계학습을 통한 수문인자 예측기술은 권역별 수계별 홍수 및 가뭄대응 계획을 능동적으로 수립하는데 활용될 것으로 판단된다. 향후 연구에서는 LSTM의 입력영역을 Bayesian 추론기법을 활용하여 구성함으로 학습과정의 불확실성을 정량적으로 제어하고자 한다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.150-150
• /
• 2017

기후변화로 인하여 극한 홍수와 극한 가뭄 발생이 증가할 것으로 전망하고 있어 이에 대한 위험이 대두되고 있는 실정이다. 홍수 및 가뭄 수문시계열의 빈도해석시에 일반적으로 활용되는 정상성 빈도해석기법은 수문자료의 정상성을 기반으로 한 빈도해석이 대부분이기 때문에 기후변화 및 수문자료의 비정상성을 반영한 새로운 빈도해석 기법이 요구되고 있는 상황이다. 본 연구에서는 5개의 대표 관측지점(서울, 포항, 추풍령, 여수, 광주)를 선별하고 1976년부터 2015년까지 일강우자료를 활용하여 기상학적 가뭄지수인 SPI(Standardized Precipitation Index)를 산정하였다. 산정한 SPI의 경향성을 Mann-Kendall 분석을 하였으며, 정상성 및 비정상성 빈도해석을 위하여 최적확률분포로 선정된 GEV 분포 적용하였다. 본 연구에서는 가뭄빈도해석을 위하여 SPI를 입력자료로 활용하였으며, 산정된 SPI의 비정상성을 반영한 비정상성 빈도해석의 경우 Bayesian 모형을 기반으로 한 MCMC(Markov Chain Monte Carlo) 모의를 이용하여 극치분포의 사후분포 매개변수를 추정하였다. 추정 값을 바탕으로 하여 가뭄의 관측소별 빈도해석을 실시하였고 재현기간별-지속기간별 가뭄심도를 추정하여 관측소별 가뭄심도-지속기간-빈도(SDF,Severity-Duration-Frequency) 곡선을 유도하였다. 본 연구를 통하여 정상성과 비정상성 빈도해석 결과의 비교연구를 수행하였으며 기후변화에 따른 비정상 시계열로 구성된 가뭄빈도해석에 매우 유용하게 적용될 수 있을 것으로 나타났다.

• Journal of Korean Society of Transportation
• /
• v.18 no.4
• /
• pp.117-126
• /
• 2000

Conventional methods for collecting origin-destination trips have been mainly relied on the surveys of home or roadside interview. However, the methods tend to be costly, labor intensive and time disruptive to the trip makers, thus the methods are not considered suitable for Planning applications such as routing guidance, arterial management and information Provision, as the parts of deployments in Intelligent Transport Systems Motivated by the problems, more economic ways to estimate origin-destination trip tables have been studied since the late 1970s. Some of them, which have been estimating O-D table from link traffic counts are generally Entropy maximizing, Maximum likelihood, Generalized least squares(GLS), and Bayesian inference estimation etc. In the Paper, with user equilibrium constraint we formulate GLS problem for estimating O-D trips and develop a solution a1gorithm by using Genetic Algorithm, which has been known as a g1oba1 searching technique. For the purpose of evaluating the method, we apply it to Seoul inner ringroad and compare it with gradient method proposed by Spiess(1990). From the resu1ts we fond that the method developed in the Paper is superior to other.

• Journal of Korea Water Resources Association
• /
• v.50 no.3
• /
• pp.139-153
• /
• 2017

The variation in runoff due to global climate change and urbanization should be identified quantitatively because these two factors have been significantly accelerated during the last three decades in South Korea. However, only a few research to analyze the impacts due to two factors over different time scales can be found. Therefore, in this study, the hydrological model based approach and the hydrological sensitivity approach were used to separate relative impacts by two factors on monthly, seasonal, and annual time scales at the Soyang River upper basin and the Seom River basin in South Korea. The 3 techniques such as the double mass curve method, the Pettitt's test, and the BCP analysis were performed to detect change point occurred by abrupt change in the collected observed runoff. After detection of change ponts, SWAT models calibrated on the natural periods were used to calculate the changes due to human activities. Also, 6 Budyko based methods were auxiliary to verify the results from hydrological based approach.

• Asia Marketing Journal
• /
• v.11 no.1
• /
• pp.73-92
• /
• 2009

Consumers influence other consumers' brand choice behavior by delivering a variety of objective or subjective information on a particular product, which is called WOM (Word-Of-Mouth) activities. For WOM activities, WOM senders should choose messages to deliver to other consumers. We classify the contents of the messages a consumer chooses for WOM delivery into two categories: Subjective (positive or negative) evaluation and objective information on products. In our study, we regard WOM senders' activities as a choice behavior and introduce a choice model to study the relationship between the choice of different WOM information (WOM with positive or negative subjective evaluation and WOM with objective information) and its influencing factors (information sources and consumer characteristics) by developing two bivariate Probit models. In order to consider the mediating effects of WOM senders' product involvement, product attitude, and their characteristics (gender and age), we develop three second-level models for the propagation of positive evaluations, of negative evaluations, and of objective information on products in an hierarchical Bayesian modeling framework. Our empirical results show that WOM senders' information choice behavior differs according to the types of information sources. The effects of information sources on WOM activities differ according to the types of WOM messages (subjective evaluation (positive or negative) and objective information). Therefore, our study concludes that WOM activities can be partially managed with effective communication plans influencing on consumers' WOM message choice behavior. The empirical results provide some guidelines for consumers' propagation of information on products companies want.

• Journal of Intelligence and Information Systems
• /
• v.25 no.2
• /
• pp.39-55
• /
• 2019

Recently banks and large financial institutions have introduced lots of Robo-Advisor products. Robo-Advisor is a Robot to produce the optimal asset allocation portfolio for investors by using the financial engineering algorithms without any human intervention. Since the first introduction in Wall Street in 2008, the market size has grown to 60 billion dollars and is expected to expand to 2,000 billion dollars by 2020. Since Robo-Advisor algorithms suggest asset allocation output to investors, mathematical or statistical asset allocation strategies are applied. Mean variance optimization model developed by Markowitz is the typical asset allocation model. The model is a simple but quite intuitive portfolio strategy. For example, assets are allocated in order to minimize the risk on the portfolio while maximizing the expected return on the portfolio using optimization techniques. Despite its theoretical background, both academics and practitioners find that the standard mean variance optimization portfolio is very sensitive to the expected returns calculated by past price data. Corner solutions are often found to be allocated only to a few assets. The Black-Litterman Optimization model overcomes these problems by choosing a neutral Capital Asset Pricing Model equilibrium point. Implied equilibrium returns of each asset are derived from equilibrium market portfolio through reverse optimization. The Black-Litterman model uses a Bayesian approach to combine the subjective views on the price forecast of one or more assets with implied equilibrium returns, resulting a new estimates of risk and expected returns. These new estimates can produce optimal portfolio by the well-known Markowitz mean-variance optimization algorithm. If the investor does not have any views on his asset classes, the Black-Litterman optimization model produce the same portfolio as the market portfolio. What if the subjective views are incorrect? A survey on reports of stocks performance recommended by securities analysts show very poor results. Therefore the incorrect views combined with implied equilibrium returns may produce very poor portfolio output to the Black-Litterman model users. This paper suggests an objective investor views model based on Support Vector Machines(SVM), which have showed good performance results in stock price forecasting. SVM is a discriminative classifier defined by a separating hyper plane. The linear, radial basis and polynomial kernel functions are used to learn the hyper planes. Input variables for the SVM are returns, standard deviations, Stochastics %K and price parity degree for each asset class. SVM output returns expected stock price movements and their probabilities, which are used as input variables in the intelligent views model. The stock price movements are categorized by three phases; down, neutral and up. The expected stock returns make P matrix and their probability results are used in Q matrix. Implied equilibrium returns vector is combined with the intelligent views matrix, resulting the Black-Litterman optimal portfolio. For comparisons, Markowitz mean-variance optimization model and risk parity model are used. The value weighted market portfolio and equal weighted market portfolio are used as benchmark indexes. We collect the 8 KOSPI 200 sector indexes from January 2008 to December 2018 including 132 monthly index values. Training period is from 2008 to 2015 and testing period is from 2016 to 2018. Our suggested intelligent view model combined with implied equilibrium returns produced the optimal Black-Litterman portfolio. The out of sample period portfolio showed better performance compared with the well-known Markowitz mean-variance optimization portfolio, risk parity portfolio and market portfolio. The total return from 3 year-period Black-Litterman portfolio records 6.4%, which is the highest value. The maximum draw down is -20.8%, which is also the lowest value. Sharpe Ratio shows the highest value, 0.17. It measures the return to risk ratio. Overall, our suggested view model shows the possibility of replacing subjective analysts's views with objective view model for practitioners to apply the Robo-Advisor asset allocation algorithms in the real trading fields.

• Journal of the Korean Data and Information Science Society
• /
• v.27 no.3
• /
• pp.587-598
• /
• 2016

In predicting an outcome of election using a variety of methods ahead of the election, non-response is one of the major issues. Therefore, to address the non-response issue, a variety of methods of non-response imputation may be employed, but the result of forecasting tend to vary according to methods. In this study, in order to improve electoral forecasts, we studied a model based method of non-response imputation attempting to apply the Monte Carlo Expectation Maximization (MCEM) algorithm, introduced by Wei and Tanner (1990). The MCEM algorithm using maximum likelihood estimates (MLEs) is applied to solve the boundary solution problem under the non-ignorable non-response mechanism. We performed the simulation studies to compare estimation performance among MCEM, maximum likelihood estimation, and Bayesian estimation method. The results of simulation studies showed that MCEM method can be a reasonable candidate for non-response model estimation. We also applied MCEM method to the Korean presidential election exit poll data of 2012 and investigated prediction performance using modified within precinct error (MWPE) criterion (Bautista et al., 2007).

• Proceedings of the Korea Technology Innovation Society Conference
• /
• 2003.11a
• /
• pp.67-85
• /
• 2003

본 논문에서는 최근 통신 산업에서 중요한 쟁점으로 떠오르고 있는 단말기 무선이동 통신 단말기의 발전 방향을 소비자 선호에 기반하여 살펴보았다. 소비자 선호 정보를 얻기 위하여 컨조인트(conjoint) 분석 방법이 사용되었다. 컨조인트 방법은 가상의 대안들에 대한 응답자의 진술 선호에 기반을 두기 때문에 미래의 무선 이동통신 단말기에 대해 분석하는데 적합한 방법이다. 컨조인트 방법을 위한 설문은 대한민국 서울에서 445 명의 성인남녀를 대상으로 행해졌다. 소비자의 이질적인 선호를 알기 위해 혼합 로짓(mixed legit) 모형을 사용하였다. 추정은 최근 새로운 시뮬레이션 기법으로 떠오르고 있는 베이지안(Bayesian) 방법을 이용하였다. 선호의 분포 가정으로 기존의 일관적인 정규 분포 가정과 달리 가격 계수를 위하여 로그 정규(lognormal) 분포, 하이퀄리티 인터넷 특성과 PC 프로그램 호환 가능 여부의 계수들에 대해서 잘린 정규(censored normal) 분포를 가정 하였다. 추정 결과 무선 이동 통신 단말기의 각 속성들에 대한 응답자들간 선호가 크게 차이 나는 것을 알 수 있었다. 화면 크기의 경우에는 대부분의 소비자들이 현재 일반적인 핸드폰보다는 큰 화면을 선호한다는 것과 휴대성을 상당히 고려한다는 것을 간접적으로 알 수 있었다. 또한, 소비자들이 무선 이동 통신 단말기가 휴대 인터넷과 PC 프로그램 호환이 가능한지 여부에는 대부분 무관심하다는 것을 알 수 있었다. 본 논문의 결과는 차세대 무선 이동 통신 단말기의 속성 조합 시 고려해야 할 점과 휴대 인터넷 서비스의 방향에 대해 시사점을 줄 수 있을 것이다.각 73.44±0.87%, 72.88±0.25%의 함량이였다. 운동사육시간이 길어질수록 운동사육구에서는 수분함량이 운동5일째에는 73.56±0.22%였으며, 운동 20일에는 75.88±0.94%로 초기수분함량보다는 3%정도 증가하였다. 반면, 비운동사육구에서는 큰 변화를 나타내고 있지 않았다(p<0.05). 운동과 비운동시킨 참돔의 지질 함량의 변화는 운동시킨 참돔은 운동으로 인한 에너지 소비로 인하여 함량이 유의적으로 감소했으며(r=-0.35), 비운동사육구에서는 절식으로 인하여 지질함량이 감소하였다(r=-0.38). 파괴강도와 가장 밀접한 영향을 가지는 콜라겐은 운동과 비운동 모두 사육기간동안 큰 변화는 보이지 않았다. 초기의 파괴강도값은 1.45±0.02kg(운동사육구), 1.36±0.18kg(비운동사육구)이였으며 사육기간동안 운동사육구는 파괴강도값이 증가한 반면, 비운동수조에서는 참돔의 파괴강도는 사육기간동안 큰 유의차가 없었다. 각 성분간의 상관도를 살펴보면, 수분함량과 파괴강도는 상관성을 가졌으며, 지질함량과 파괴강도도 같은 경향은 나타내었다. 운동기간동안의 파괴강도와 콜라겐 사이에는 상관성의 거의 없었다. 이는 운동기간에 따른 파괴강도의 증가가 콜라겐의 함량의 증가보다는 지질함량의 감소와 수분함량의 증가와 같은 성분과의 상관성이 크다고 판단된다. 다음으로는, 운동횟수에 의한 영향으로써 운동시간을 1일 6시간으로 설정하여, 운동횟수를 결정하기 위하여 오전, 오후에 각 3시간씩 운동시키는 방법과 오전부터 6시간동안 운동시키는 두 방법을 이용하여 품질을 비교하였다. 각 조건에 따라 운동시킨 참돔의 수분함량을 나타낸 것으로, 2회(오전

• Korean journal of applied entomology
• /
• v.51 no.4
• /
• pp.431-438
• /
• 2012

The developmental time of the nymphs of Myzus persicae was studied in the laboratory (six constant temperatures from 15 to $30^{\circ}C$ with 50~60% RH, and a photoperiod of 14L:10D) and in a green-pepper plastic house. Mortality of M. persicae in laboratory was high in the first(6.7~13.3%) and second instar nymphs(6.7%) at low temperatures and high in the third (17.8%) and fourth instar nymphs(17.8%) at high temperatures. Mortality was 66.7% at $33^{\circ}C$ in laboratory and $26.7^{\circ}C$ in plastic house. The total developmental time was the longest at $14.6^{\circ}C$ (14.4 days) and shortest at $26.7^{\circ}C$ (6.0 days) in plastic house. The lower threshold temperature of the total nymphal stage was $3.0^{\circ}C$ in laboratory. The thermal constant required for nymphal stage was 111.1DD. The relationship between developmental rate and temperature was fitted nonlinear model by Logan-6 which has the lowest value on Akaike information criterion (AIC) and Bayesian information criterion (BIC). The distribution of completion of each developmental stage was well described by the 3-parameter Weibull function ($r^2=0.95{\sim}0.97$). This model accurately described the predicted and observed occurrences. Thus the model is considered to be good for use in predicting the optimal spray time for Myzus persicae.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.20 no.4
• /
• pp.89-101
• /
• 2017

Crime occurs differently based on not only place locations and building uses but also the characteristics of the people who use the place and the spatial structures of the buildings and locations. Therefore, if spatial big data, which contain spatial and regional properties, can be utilized, proper crime prevention measures can be enacted. Recently, with the advent of big data and the revolutionary intelligent information era, predictive policing has emerged as a new paradigm for police activities. Based on 7420 actual crime incidents occurring over three years in a typical provincial city, "J city," this study identified the areas in which crimes occurred and predicted risky areas. Spatial regression analysis was performed using spatial big data about only physical and environmental variables. Based on the results, using the street width, average number of building floors, building coverage ratio, the type of use of the first floor (Type II neighborhood living facility, commercial facility, pleasure use, or residential use), this study established a Crime Incident Prediction Model (CIPM) based on Bayesian probability theory. As a result, it was found that the model was suitable for crime prediction because the overlap analysis with the actual crime areas and the receiver operating characteristic curve (Roc curve), which evaluated the accuracy of the model, showed an area under the curve (AUC) value of 0.8. It was also found that a block where the commercial and entertainment facilities were concentrated, a block where the number of building floors is high, and a block where the commercial, entertainment, residential facilities are mixed are high-risk areas. This study provides a meaningful step forward to the development of a crime prediction model, unlike previous studies that explored the spatial distribution of crime and the factors influencing crime occurrence.