• 정보관리연구
• /
• 제25권1호
• /
• pp.63-79
• /
• 1994

일본(日本)의 경제(經濟) 산업(産業) 분야의 분석(分析) 시스템 및 데이터베이스 역사는 장구하나, 전문성(專門性)이 강하여 이용자(利用者)가 한정되었다. 시스템 서비스가 본격화 된 것도 얼마되지 않는다. 본고(本稿)에서는 이 시스템 서비스의 대표적인 것을 해설하였다. 활용예(活用例)에 관해선 금리예측(金利豫測)을 예로 하여 도입(導入) 활용(活用) 상(上)의 유의점에 관해서도 설명하였다. 그밖에 신문정보(新開情報)의 확인에서 보다 고도의 것으로서 경제예측(經濟豫測)이나, 경제예측(經濟豫測)을 기반으로 한 각종 모델의 이용과 산업분석(産業分析)에 관해서도 다루었다. 조사(調査)나 연구(硏究)는 상호작용의 부분이 많아서 평상시부터 시스템에 익숙해야 하며, 이용자끼리 정보를 교환하고, 기회가 있을 때마다 헬프데스크를 활용하는 등 꾸준한 노력이 필요하다.

• 한국정보통신학회논문지
• /
• 제22권10호
• /
• pp.1314-1319
• /
• 2018

현재 우리 사회는 아동의 성장발달에 대한 관심이 증가한데 비해 대한성장의학회에서 사용되고 있는 TW3 기반의 최대신장예측 기법은 수동으로 이루어지고 있어 주관적이며, 진료에 있어 다소 많은 시간과 노력을 필요로 한다는 단점이 있다. 또한 현재 딥러닝, 특히 컨볼루션 신경망을 활용해 영상을 분류하는 기술은 인간의 눈보다 더 정확한 수준으로 다양한 분야에 활용되고 있으며 의료분야 또한 예외는 아니다. 따라서 성장 예측의 신뢰도를 높이고, 진단자의 편의성을 증대하기 위해 본 논문에서는 컨볼루션 신경망을 이용해 좌측 수골의 발달 수준을 예측하고 소아청소년의 최대신장예측에 활용되는 딥러닝을 이용한 TW3 알고리즘을 제안한다.

• 한국콘텐츠학회논문지
• /
• 제21권6호
• /
• pp.552-560
• /
• 2021

새로운 미술품 유통방식의 발달로 미술품의 미적 효용을 넘어 투자재로서 바라보는 시각이 활성화되고 있다. 미술품의 가격은 주식이나 채권 등과 달리 객관적 요소와 주관적 요소들이 모두 반영되어 결정되는 이질적 특성이 있기 때문에 가격 예측에 있어서 그 불확실성이 높다. 본 연구에서는 LSTM(장단기 기억) 순환신경망 딥러닝 모형을 활용하여 낙찰총액 순위 1위부터 10위까지의 한국 작가의 회화 작품을 대상으로 작가의 특성, 작품의 물리적 특성, 판매적 특성 등을 입력으로 하여 경매 낙찰가의 예측을 시도하였다. 연구 결과, 모델에 의한 예측 가격과 실제 낙찰 가격의 차이를 설명하는 RMSE 값이 0.064 수준이었으며 작가별로는 이대원 작가의 예측력이 가장 높았고, 이중섭 작가의 예측력이 가장 낮았다. 투자재로서 미술품 시장이 더욱 활성화되고 경매 낙찰 가격의 예측 수요가 높아지면서 본 연구의 결과가 활용될 수 있을 것이다.

• 한국항공우주학회지
• /
• 제50권3호
• /
• pp.215-222
• /
• 2022

항공교통흐름관리의 목적은 공항 및 공역의 수용량 안에서 항공교통 수요를 만족시키는 것이다. 그러므로 수용량을 정확하게 예측하는 것은 항공교통흐름관리의 성능에 많은 영향을 준다. 본 논문은 특정 공항의 예상 출·도착 수요, 시각, 기상 및 실제 처리한 항공기 대수 등 과거의 항공기운항 데이터를 기계학습의 한 방법론인 부스팅 앙상블 알고리즘으로 학습하여 시간당 출·도착하는 항공기의 수를 예측하는 회귀모형을 개발하였다. 기계학습을 통해 도출된 모델은 실제 인천국제공항의 출·도착 항공편 데이터를 이용해 검증하였으며, 결정계수가 0.95 이상으로 나타났다. 이 모델을 이용하여 접근관제구역의 수용량을 간접적으로 예측할 수 있었다.

• 지능정보연구
• /
• 제28권4호
• /
• pp.81-94
• /
• 2022

최근 고액의 실물자산이나 채권을 분할하여 여러 투자자가 공동으로 투자하는 이른바 조각투자가 인기를 얻고 있다. 2016년 설립된 뮤직카우는 음원 유통에 따른 저작권료 참여 청구권을 조각투자할 수 있는 서비스를 세계 최초로 시작하였다. 본 연구에서는 딥러닝 알고리즘 중 하나인 LSTM 모델을 사용하여 뮤직카우에서 거래되는 저작권료 참여 청구권의 가격을 예측하는 연구를 진행하였다. 청구권의 이전 가격과 거래량, 저작권료와 같은 청구권과 관련된 변수 외에도, 음악저작권료 참여 청구권 시장 상황을 나타내는 종합 지표와 경제 상황을 반영하는 환율, 국고채 금리, 한국종합주가지수도 변수로 사용하였다. 연구 결과 상대적으로 거래량이 낮은 조각투자의 사례에서도 LSTM 모델이 거래가격을 잘 예측하는 것을 확인할 수 있었다.

• 산업경영시스템학회지
• /
• 제46권2호
• /
• pp.102-108
• /
• 2023

Recently in Korea, YouTube stock channels increased rapidly due to the high social interest in the stock market during the COVID-19 period. Accordingly, the role of new media channels such as YouTube is attracting attention in the process of generating and disseminating market information. Nevertheless, prior studies on the market forecasting power of YouTube stock channels remain insignificant. In this study, the market forecasting power of the information from the YouTube stock channel was examined and compared with traditional news media. To measure information from each YouTube stock channel and news media, positive and negative opinions were extracted. As a result of the analysis, opinion in channels operated by media outlets were found to be leading indicators of KOSPI market returns among YouTube stock channels. The prediction accuracy by using logistic regression model show 74%. On the other hand, Sampro TV, a popular YouTube stock channel, and the traditional news media simply reported the market situation of the day or instead showed a tendency to lag behind the market. This study is differentiated from previous studies in that it verified the market predictive power of the information provided by the YouTube stock channel, which has recently shown a growing trend in Korea. In the future, the results of advanced analysis can be confirmed by expanding the research results for individual stocks.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2021년도 학술발표회
• /
• pp.131-131
• /
• 2021

Deep learning models, especially those based on long short-term memory (LSTM), have presented their superiority in addressing time series data issues recently. This study aims to comprehensively evaluate the performance of deep learning models that belong to the supervised learning category in streamflow prediction. Therefore, six deep learning models-standard LSTM, standard gated recurrent unit (GRU), stacked LSTM, bidirectional LSTM (BiLSTM), feed-forward neural network (FFNN), and convolutional neural network (CNN) models-were of interest in this study. The Red River system, one of the largest river basins in Vietnam, was adopted as a case study. In addition, deep learning models were designed to forecast flowrate for one- and two-day ahead at Son Tay hydrological station on the Red River using a series of observed flowrate data at seven hydrological stations on three major river branches of the Red River system-Thao River, Da River, and Lo River-as the input data for training, validation, and testing. The comparison results have indicated that the four LSTM-based models exhibit significantly better performance and maintain stability than the FFNN and CNN models. Moreover, LSTM-based models may reach impressive predictions even in the presence of upstream reservoirs and dams. In the case of the stacked LSTM and BiLSTM models, the complexity of these models is not accompanied by performance improvement because their respective performance is not higher than the two standard models (LSTM and GRU). As a result, we realized that in the context of hydrological forecasting problems, simple architectural models such as LSTM and GRU (with one hidden layer) are sufficient to produce highly reliable forecasts while minimizing computation time because of the sequential data nature.

• 산업경영시스템학회지
• /
• 제46권1호
• /
• pp.23-31
• /
• 2023

Research and interest in sustainable printing are increasing in the packaging printing industry. Currently, predicting the amount of ink required for each work is based on the experience and intuition of field workers. Suppose the amount of ink produced is more than necessary. In this case, the rest of the ink cannot be reused and is discarded, adversely affecting the company's productivity and environment. Nowadays, machine learning models can be used to figure out this problem. This study compares the ink usage prediction machine learning models. A simple linear regression model, Multiple Regression Analysis, cannot reflect the nonlinear relationship between the variables required for packaging printing, so there is a limit to accurately predicting the amount of ink needed. This study has established various prediction models which are based on CART (Classification and Regression Tree), such as Decision Tree, Random Forest, Gradient Boosting Machine, and XGBoost. The accuracy of the models is determined by the K-fold cross-validation. Error metrics such as root mean squared error, mean absolute error, and R-squared are employed to evaluate estimation models' correctness. Among these models, XGBoost model has the highest prediction accuracy and can reduce 2134 (g) of wasted ink for each work. Thus, this study motivates machine learning's potential to help advance productivity and protect the environment.

• 한국환경보건학회지
• /
• 제50권2호
• /
• pp.113-124
• /
• 2024

Background: The increasing need to minimize animal testing has sparked interest in alternative methods with more humane, cost-effective, and time-saving attributes. In particular, in silico-based computational toxicology is gaining prominence. Adverse outcome pathway (AOP) is a biological map depicting toxicological mechanisms, composed of molecular initiating events (MIEs), key events (KEs), and adverse outcomes (AOs). To understand toxicological mechanisms, predictive models are essential for AOP components in computational toxicology, including molecular structures. Objectives: This study reviewed the literature and investigated previous research cases related to AOP and in silico methodologies. We describe the results obtained from the analysis, including predictive techniques and approaches that can be used for future in silico-based alternative methods to animal testing using AOP. Methods: We analyzed in silico methods and databases used in the literature to identify trends in research on in silico prediction models. Results: We reviewed 26 studies related to AOP and in silico methodologies. The ToxCast/Tox21 database was commonly used for toxicity studies, and MIE was the most frequently used predictive factor among the AOP components. Machine learning was most widely used among prediction techniques, and various in silico methods, such as deep learning, molecular docking, and molecular dynamics, were also utilized. Conclusions: We analyzed the current research trends regarding in silico-based alternative methods for animal testing using AOPs. Developing predictive techniques that reflect toxicological mechanisms will be essential to replace animal testing with in silico methods. In the future, since the applicability of various predictive techniques is increasing, it will be necessary to continue monitoring the trend of predictive techniques and in silico-based approaches.

• 대기
• /
• 제34권2호
• /
• pp.177-185
• /
• 2024

As increasing global interest in renewable energy due to the ongoing climate crisis, there is a growing need for efficient technologies to manage such resources. This study focuses on the predictive skill of daily solar power generation using weather observation and forecast data. Meteorological data from the Korea Meteorological Administration and solar power generation data from the Korea Power Exchange were utilized for the period from January 2017 to May 2023, considering both inland (Daejeon) and coastal (Busan) regions. Temperature, wind speed, relative humidity, and precipitation were selected as relevant meteorological variables for solar power prediction. All data was preprocessed by removing their systematic components to use only their residuals and the residual of solar data were further processed with weighted adjustments for homoscedasticity. Four models, MLR (Multiple Linear Regression), RF (Random Forest), DNN (Deep Neural Network), and RNN (Recurrent Neural Network), were employed for solar power prediction and their performances were evaluated based on predicted values utilizing observed meteorological data (used as a reference), 1-day-ahead forecast data (referred to as fore1), and 2-day-ahead forecast data (fore2). DNN-based prediction model exhibits superior performance in both regions, with RNN performing the least effectively. However, MLR and RF demonstrate competitive performance comparable to DNN. The disparities in the performance of the four different models are less pronounced than anticipated, underscoring the pivotal role of fitting models using residuals. This emphasizes that the utilized preprocessing approach, specifically leveraging residuals, is poised to play a crucial role in the future of solar power generation forecasting.