• 제목/요약/키워드: absolute model accuracy

검색결과 255건 처리시간 0.022초

정밀 GPS 위성궤도 결정 및 오차 특성 분석 (Precision GPS Orbit Determination and Analysis of Error Characteristics)

  • 배태석
    • 한국측량학회지
    • /
    • 제27권4호
    • /
    • pp.437-444
    • /
    • 2009
  • 동역학적 방법을 이용한 GPS(Global Positioning System) 위성궤도 결정을 위해 양방향 적분이 가능한 multi-step 방식의 수치적분기를 개발하였으며, 이는 GPS 위성 고도에서 마이크로미터 수준의 정확도를 보였다. 가속도 모델링에서 달, 태양 이외의 천체에 의한 인력은 매우 작으므로 태양복사압에서 경험적 모델로 대체하였다. 위성궤도 미지수는 수치적분된 위성궤도와 IGS(International GNSS Service) 정밀궤도를 이용하여 최소제곱방법으로 결정했다. 이를 위해서는 수치적분기에서 가속도와 함께 미지수에 대한 편미분값을 동시에 적분해야 한다. 추정된 위성궤도 미지수를 이용하여 계산한 잔차의 RMS(Root Mean Squares error)로 부터 위성궤도의 정확도를 검증했다. 2009년 3월 한달의 평균적인 궤도오차 RMS는 5.2mm 였으며, 궤도오차의 절대적인 크기는 위성체의 종류 및 위성진행방향기준 좌표계 상에서 특별히 편향된 형태를 보이지는 않는 것으로 나타났다. 본 연구에서 적용한 태양복사압 모델은 상수항 및 궤도당 1주기에 대한 변화만을 포함하고 있으므로, 궤도당 2주기에 해당하는 궤도오차 양상을 크게 보이고 있으며 이에 대한 추가적인 연구가 필요할 것으로 판단된다.

머신러닝을 통한 잉크 필요량 예측 알고리즘 (Machine Learning Algorithm for Estimating Ink Usage)

  • 권세욱;현영주;태현철
    • 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.

관개용수로 CCTV 이미지를 이용한 CNN 딥러닝 이미지 모델 적용 (Application of CCTV Image and Semantic Segmentation Model for Water Level Estimation of Irrigation Channel)

  • 김귀훈;김마가;윤푸른;방재홍;명우호;최진용;최규훈
    • 한국농공학회논문집
    • /
    • 제64권3호
    • /
    • pp.63-73
    • /
    • 2022
  • A more accurate understanding of the irrigation water supply is necessary for efficient agricultural water management. Although we measure water levels in an irrigation canal using ultrasonic water level gauges, some errors occur due to malfunctions or the surrounding environment. This study aims to apply CNN (Convolutional Neural Network) Deep-learning-based image classification and segmentation models to the irrigation canal's CCTV (Closed-Circuit Television) images. The CCTV images were acquired from the irrigation canal of the agricultural reservoir in Cheorwon-gun, Gangwon-do. We used the ResNet-50 model for the image classification model and the U-Net model for the image segmentation model. Using the Natural Breaks algorithm, we divided water level data into 2, 4, and 8 groups for image classification models. The classification models of 2, 4, and 8 groups showed the accuracy of 1.000, 0.987, and 0.634, respectively. The image segmentation model showed a Dice score of 0.998 and predicted water levels showed R2 of 0.97 and MAE (Mean Absolute Error) of 0.02 m. The image classification models can be applied to the automatic gate-controller at four divisions of water levels. Also, the image segmentation model results can be applied to the alternative measurement for ultrasonic water gauges. We expect that the results of this study can provide a more scientific and efficient approach for agricultural water management.

Construction of a Novel Mitochondria-Associated Gene Model for Assessing ESCC Immune Microenvironment and Predicting Survival

  • Xiu Wang;Zhenhu Zhang;Yamin Shi;Wenjuan Zhang;Chongyi Su;Dong Wang
    • Journal of Microbiology and Biotechnology
    • /
    • 제34권5호
    • /
    • pp.1164-1177
    • /
    • 2024
  • Esophageal squamous cell carcinoma (ESCC) is among the most common malignant tumors of the digestive tract, with the sixth highest fatality rate worldwide. The ESCC-related dataset, GSE20347, was downloaded from the Gene Expression Omnibus (GEO) database, and weighted gene co-expression network analysis was performed to identify genes that are highly correlated with ESCC. A total of 91 transcriptome expression profiles and their corresponding clinical information were obtained from The Cancer Genome Atlas database. A mitochondria-associated risk (MAR) model was constructed using the least absolute shrinkage and selection operator Cox regression analysis and validated using GSE161533. The tumor microenvironment and drug sensitivity were explored using the MAR model. Finally, in vitro experiments were performed to analyze the effects of hub genes on the proliferation and invasion abilities of ESCC cells. To confirm the predictive ability of the MAR model, we constructed a prognostic model and assessed its predictive accuracy. The MAR model revealed substantial differences in immune infiltration and tumor microenvironment characteristics between high- and low-risk populations and a substantial correlation between the risk scores and some common immunological checkpoints. AZD1332 and AZD7762 were more effective for patients in the low-risk group, whereas Entinostat, Nilotinib, Ruxolutinib, and Wnt.c59 were more effective for patients in the high-risk group. Knockdown of TYMS significantly inhibited the proliferation and invasive ability of ESCC cells in vitro. Overall, our MAR model provides stable and reliable results and may be used as a prognostic biomarker for personalized treatment of patients with ESCC.

연직수문의 퇴적토 배출특성에 관한 실험적 연구 (An Experimental Study on the Sediment Transport Characteristics Through Vertical Lift Gate)

  • 이지행;최흥식
    • Ecology and Resilient Infrastructure
    • /
    • 제5권4호
    • /
    • pp.276-284
    • /
    • 2018
  • 하단배출 형태의 연직수문을 대상으로 퇴적토 배출특성에 따른 두부침식 거리비, 퇴적토 이동거리와 이동량을 분석하기 위해 수리 모형실험과 차원해석을 수행하였다. Froude 수와 배출특성의 상관관계를 도식화하고, 퇴적토 배출특성을 지배하는 무차원 매개변수에 의한 다중회귀식을 제안하였다. 두부침식거리, 퇴적토 이동거리와 이동량에 대한 각 다중회귀 분석식의 결정계수는 각각 0.618, 0.632, 0.866으로 높게 나타났다. 개발한 퇴적토 배출특성식의 사용성을 평가하기 위해 실제 측정값과 회귀분석식에 의해 계산된 값의 95%의 예측 신뢰구간 분석을 수행하였고, 두부침식거리, 퇴적토 이동거리와 이동량에 대한 예측의 정확도 분석차원의 NSE (Nash-Sutcliffe Efficiency), RMSE (root mean square)와 MAPE (mean absolute percentage error)는 적절한 것으로 판단되었다.

Prediction of Vertical Sea Water Temperature Profile in the East Sea Based on Machine Learning and XBT Data

  • Kim, Young-Joo;Lee, Soo-Jin;Kim, Young-Won
    • 한국컴퓨터정보학회논문지
    • /
    • 제27권11호
    • /
    • pp.47-55
    • /
    • 2022
  • 최근 우리나라에서도 인공지능 모델을 이용한 수온예측 관련 연구가 활발히 진행되고 있으나 한반도 주변 해역의 수온을 예측한 대부분의 연구는 주로 해수면 온도 예측에 중점을 두고 진행되었다. 반면 본 연구는 XBT(eXpendable Bathythermograph, 소모성 연직수온측정기) 데이터와 기계 학습 모델(RandomForest, XGBoost, LightGBM)을 사용하여 잠수함 작전 및 대잠전(Anti-Submarine Warfare)에 있어서 군사적으로 중요한 동해의 수직 수온분포를 예측하였다. 동해 특정해역의 해수면부터 수심 200m까지 측정된 XBT 데이터를 이용하여 모델을 학습시키고 절대 평균 오차(MAE, Mean Absolute Error)와 수직 수온분포 그래프를 통해 예측정확도를 평가하였다.

Rational Function Model 기반 KOMPSAT-3A 스트립 번들조정 (Bundle Adjustment of KOMPSAT-3A Strip Based on Rational Function Model)

  • 윤완상;김태정
    • 대한원격탐사학회지
    • /
    • 제34권3호
    • /
    • pp.565-578
    • /
    • 2018
  • 본 연구에서는 번들조정 과정에서 요구되는 GCP의 수를 줄이기 위해 동일궤도 상의 개별 영상 대신 스트립을 스트립을 모델링 할 수 있는 가능성을 조사한다. 이를 위해 먼저 동일 궤도상에 존재하는 각 개별영상의 RFM(Rational function model)으로부터 스트립에 대한 RFM을 생성하였다. 다음으로, 생성된 스트립 이미지 간의 번들 조정을 통해 모델 보정계수를 산출하였다. 실험을 위해 각 3개의 Scene 영상으로 구성된 KOMPSAT-3A 스테레오 스트립을 사용하였다. 실험을 통해 스트립의 특정지역에 위치한 기준점만을 사용하여 초기모델 개선이 가능함을 확인하였다. 또한 12개의 지상기준점을 사용한 스트립 번들조정 수행 결과 수평 수직 방향으로 약 2 m의 3차원 위치 결정이 가능함을 확인하였다. 이를 통해 단일 영상 기반 번들조정보다 스트립 번들조정이 더 효율적일 수 있음을 확인하였다.

앙상블 칼만필터를 연계한 추계학적 연속형 저류함수모형 (II) : - 적용 및 검증 - (Stochastic Continuous Storage Function Model with Ensemble Kalman Filtering (II) : Application and Verification)

  • 이병주;배덕효
    • 한국수자원학회논문집
    • /
    • 제42권11호
    • /
    • pp.963-972
    • /
    • 2009
  • 본 연구의 목적은 앙상블 칼만필터 기법과 연속형 저류함수모형을 연계하여 개발한 추계학적 연속형 저류함수모형의 적용성을 평가하고자 하는데 있다. 대상유역은 안동댐과 임하댐을 포함하는 지보 수위관측소 상류유역을 선정하였으며 2006년과 2007년 홍수기에 대해 분석을 수행하였다. 확정론적 모형을 적용한 결과 장기간의 모의기간에 대해 유출해석이 가능한 것을 확인하였다. 앙상블 칼만필터 기법을 적용하기 위해 Monte Carlo 모의기법을 적용하여 모형입력자료와 매개변수들에 대해 앙상블 멤버를 생성하였다. 추계학적 모형과 확정론적 모형의 누적절대오차를 비교한 결과 안동댐과 임하댐의 2007년 사상에서 각각 17.5 %와 18.3 %의 정확도가 향상되고 지보수위관측소에서는 40 % 이상의 정확도가 향상되는 것으로 나타났다. 이상의 결과로부터 관측유량과의 오차가 큰 모의결과에 있어서는 추계학적 모형이 보다 향상된 결과를 도출하는 것을 확인하였다.

Estimating the unconfined compression strength of low plastic clayey soils using gene-expression programming

  • Muhammad Naqeeb Nawaz;Song-Hun Chong;Muhammad Muneeb Nawaz;Safeer Haider;Waqas Hassan;Jin-Seop Kim
    • Geomechanics and Engineering
    • /
    • 제33권1호
    • /
    • pp.1-9
    • /
    • 2023
  • The unconfined compression strength (UCS) of soils is commonly used either before or during the construction of geo-structures. In the pre-design stage, UCS as a mechanical property is obtained through a laboratory test that requires cumbersome procedures and high costs from in-situ sampling and sample preparation. As an alternative way, the empirical model established from limited testing cases is used to economically estimate the UCS. However, many parameters affecting the 1D soil compression response hinder employing the traditional statistical analysis. In this study, gene expression programming (GEP) is adopted to develop a prediction model of UCS with common affecting soil properties. A total of 79 undisturbed soil samples are collected, of which 54 samples are utilized for the generation of a predictive model and 25 samples are used to validate the proposed model. Experimental studies are conducted to measure the unconfined compression strength and basic soil index properties. A performance assessment of the prediction model is carried out using statistical checks including the correlation coefficient (R), the root mean square error (RMSE), the mean absolute error (MAE), the relatively squared error (RSE), and external criteria checks. The prediction model has achieved excellent accuracy with values of R, RMSE, MAE, and RSE of 0.98, 10.01, 7.94, and 0.03, respectively for the training data and 0.92, 19.82, 14.56, and 0.15, respectively for the testing data. From the sensitivity analysis and parametric study, the liquid limit and fine content are found to be the most sensitive parameters whereas the sand content is the least critical parameter.

Comparison of the accuracy of digitally fabricated polyurethane model and conventional gypsum model

  • Kim, So-Yeun;Lee, So-Hyoun;Cho, Seong-Keun;Jeong, Chang-Mo;Jeon, Young-Chan;Yun, Mi-Jung;Huh, Jung-Bo
    • The Journal of Advanced Prosthodontics
    • /
    • 제6권1호
    • /
    • pp.1-7
    • /
    • 2014
  • PURPOSE. The accuracy of a gypsum model (GM), which was taken using a conventional silicone impression technique, was compared with that of a polyurethane model (PM), which was taken using an iTero$^{TM}$ digital impression system. MATERIALS AND METHODS. The maxillary first molar artificial tooth was selected as the reference tooth. The GMs were fabricated through a silicone impression of a reference tooth, and PMs were fabricated by a digital impression (n=9, in each group). The reference tooth and experimental models were scanned using a 3 shape convince$^{TM}$ scan system. Each GM and PM image was superimposed on the registered reference model (RM) and 2D images were obtained. The discrepancies of the points registered on the superimposed images were measured and defined as GM-RM group and PM-RM group. Statistical analysis was performed using a Student's T-test (${\alpha}=0.05$). RESULTS. A comparison of the absolute value of the discrepancy revealed a significant difference between the two groups only at the occlusal surface. The GM group showed a smaller mean discrepancy than the PM group. Significant differences in the GM-RM group and PM-RM group were observed in the margins (point a and f), mesial mid-axial wall (point b) and occlusal surfaces (point c and d). CONCLUSION. Under the conditions examined, the digitally fabricated polyurethane model showed a tendency for a reduced size in the margin than the reference tooth. The conventional gypsum model showed a smaller discrepancy on the occlusal surface than the polyurethane model.