• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.6
• /
• pp.259-266
• /
• 2023

It is very difficult to measure the performance of the machine learning model in the business service stage. Therefore, managing the performance of the model through the operational department is not done effectively. Academically, various studies have been conducted on the concept drift detection method to determine whether the model status is appropriate. The operational department wants to know quantitatively the performance of the operating model, but concept drift can only detect the state of the model in relation to the data, it cannot estimate the quantitative performance of the model. In this study, we propose a performance prediction model (PPM) that quantitatively estimates precision through the statistics of concept drift. The proposed model induces artificial drift in the sampling data extracted from the training data, measures the precision of the sampling data, creates a dataset of drift and precision, and learns it. Then, the difference between the actual precision and the predicted precision is compared through the test data to correct the error of the performance prediction model. The proposed PPM was applied to two models, a loan underwriting model and a credit card fraud detection model that can be used in real business. It was confirmed that the precision was effectively predicted.

• Journal of Korean Port Research
• /
• v.12 no.2
• /
• pp.373-384
• /
• 1998

In search and rescue mission the leeway formula based on the field experiments are utilized for the estimation of wind effect on distressed targets. This paper summarized the leeway formula from the available references. In the summary the environmental data collection method and experimental conditions are described along with the formula. Also the formula currently used in CASP of the U.S. Coast Gurard and CANSARP of the Canadian Coast Guard are discussed.

• Journal of Korean Port Research
• /
• v.14 no.3
• /
• pp.353-360
• /
• 2000

In this study a drift prediction model based on fluid dynamics theory is introduced. The essential effects of environmental loads and target characteristics are taken into account from a fluid dynamics point of view. The governing equations of motion are derived from Netwon's law of dynamics. In the mathematical formulation only three degrees of freedom(surge, sway, yaw) of the drifting object are assumed and the environmental loads considered are the forces and moments by wind and current. A computer algorithm for this model is implemented to obtain the numerical result in the time domain. The preliminary tests for model verification are conducted and the results are compared with the field experiment data as well as leeway formula suggested from the field test data.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2000.11a
• /
• pp.63-71
• /
• 2000

In this study a drift prediction model based on fluid dynamics theory is introduced. The essential effects of environmental loads and target characteristics are taken into account from a fluid dynamics point of view. The governing equations of motion are derived from Newton's law of dynamics. In the mathematical formulation only three degrees of freedom(surge, sway, yaw) of the drifting object are assumed and the environmental loads considered are the forces and moments by wind and current. A computer algorithm for this model is implemented to obtain the numerical result in the time domain. The preliminary tests for model verification are conducted and the results are compared with the field experiment data as well as leeway formula suggested from the field test data.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.6 no.1
• /
• pp.111-120
• /
• 2000

In this study the drift prediction models currently used in search and rescue(SAR) operations are introduced and the basic differences among them are analyzed by comparing the target classification, the treatment of environmental conditions, leeway equations and the determination of search area. In addition the new methods under development including AP98 of U.S. Coast Guard are compared and studied to improve the efficiency of SAR operation.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2002.11a
• /
• pp.1-6
• /
• 2002

To provide an easy operation of drift prediction model in SAR(search and rescue) mission a GUI program running on Window environment has developed. Users can make choice of input data on the screen by just clicking the mouse and the prediction results of datum points and trajectories of vessels are drawn on the map. The program contains both Leeway Equation model and mathematical model. The FORTRAN language was used in programming and Lehay Winteracter 4.0 software was utilized for graphic presentation. The result of May, 2001 Busan field experiment was plotted with that of model prediction for demonstration purpose.

• Journal of Navigation and Port Research
• /
• v.26 no.5
• /
• pp.491-495
• /
• 2002

To provide an easy operation of drift prediction model in SAR(search and rescue) mission a GUI program running on Windows environment has developed. Users can make choice of input data on the screen by just clicking the mouse and the prediction results of datum points and trajectories of vessels are drawn on the electric chart. The program contains both Leeway Equation model and Mathematical model. The FORTRAN language was used in programming and Lehay Winteraction 4.0 software was utilized for graphic presentation. The result of May, 2001 Busan field experiment was plotted with that of model prediction for demonstration purpose.

• Journal of KIISE
• /
• v.43 no.6
• /
• pp.718-723
• /
• 2016

In streaming data analysis, detecting concept drift accurately is important to maintain the performance of classification model. Error rates are usually used for concept drift detection. However, by describing prediction results with only binary values of 0 or 1, useful information about a behavior pattern of a classifier can be lost. In this paper, we propose an effective concept drift detection method which describes performance pattern of a classifier by utilizing probability estimates for class prediction and detects a significant change in a classifier behavior. Experimental results on synthetic and real streaming data show the efficiency of the proposed method for detecting the occurrence of concept drift.

• International Journal of Naval Architecture and Ocean Engineering
• /
• v.10 no.1
• /
• pp.69-84
• /
• 2018

In order to analyze the characteristics of propeller exciting force, the hybrid grid is adopted and the numerical prediction of KCS ship model is performed for hull-propeller-rudder system by Reynolds-Averaged Navier Stokes (RANS) method and volume of fluid (VOF) model. Firstly, the numerical simulation of hydrodynamics for bare hull at oblique state is carried out. The results show that with the increasing of the drift angle, the coefficients of resistance, side force and yaw moment are constantly increasing, and the bigger the drift angle, the worse the overall uniformity of propeller disk. Then, propeller bearing force for hull-propeller-rudder system in oblique flow is calculated. It is found that the propeller thrust and torque fluctuation coefficient peak in drift angle are greater than that in straight line navigation, and the negative drift angle is greater than the positive. The fluctuation peak variation law of coefficient of side force and bending moment are different due to various causes.

• Structural Engineering and Mechanics
• /
• v.54 no.4
• /
• pp.741-754
• /
• 2015

This study evaluates prediction models for three EDPs (engineering demand parameters) using data from three symmetrical structures with RC walls designed according to the currently enforced Romanian seismic design code P100-1/2013. The three analyzed EDPs are: the maximum interstorey drift, the maximum top displacement and the maximum shear force at the base of the RC walls. The strong ground motions used in this study consist of three pairs of recordings from the Vrancea intermediate-depth earthquakes of 1977, 1986 and 1990, as well as two other pairs of recordings from significant earthquakes in Turkey and Greece (Erzincan and Aigion). The five pairs of recordings are rotated in a clockwise direction and the values of the EDPs are recorded. Finally, the relation between various IMs (intensity measures) of the strong ground motion records and the EDPs is studied and two prediction models for EDPs are also evaluated using the analysis of residuals.