• International Journal of Internet, Broadcasting and Communication
• /
• v.11 no.1
• /
• pp.47-53
• /
• 2019

The idea of ensemble learning is to train multiple models, each with the objective to predict or classify a set of results. Most of the errors from a model's learning are from three main factors: variance, noise, and bias. By using ensemble methods, we're able to increase the stability of the final model and reduce the errors mentioned previously. By combining many models, we're able to reduce the variance, even when they are individually not great. In this paper we propose an ensemble model and applied it to classification problem. In iris, Pima indian diabeit and semiconductor fault detection problem, proposed model classifies well compared to traditional single classifier that is logistic regression, SVM and random forest.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.167.4-167
• /
• 2001

Accurate mathematical models of spacecraft components are an essential of spacecraft attitude control system design, analysis and simulation. Gyro is one of the most important spacecraft components used for attitude propagation and control. Gyro errors may seriously degrade the accuracy of the calculated spacecraft angular rate and of attitude estimates due to inherent drift and bias errors. In this paper, a detailed mathematical model of gyro containing the relationships for predicting spacecraft angular rate and disturbances is proposed. Stochastic model describing random drift behavior is discussed in frequency domain and time domain. In order to illustrate this approach, we analyze the behavior for Litton´s Space Inertial Reference Uint(SIRU).

• Proceedings of the Korean Institute of IIIuminating and Electrical Installation Engineers Conference
• /
• 2007.11a
• /
• pp.169-171
• /
• 2007

We need two operation modes to obtain the complex hologram without bias and the conjugate image in the modified triangular interferometer(MTI). To solve the problem, we proposed the optimized MTI with one wave plate, which can obtain cosine and sine functions by the combination of one wave plate and one linear polarizer. In the extraction of phase term using the combination of polarization components, the phase error occurs, and we simulated such potential phase errors in the optimized MTI.

• Proceedings of the Korean Institute of IIIuminating and Electrical Installation Engineers Conference
• /
• 2007.05a
• /
• pp.117-119
• /
• 2007

We need two operation modes to obtain the complex hologram without bias and the conjugate image in the modified triangular interferometer(MTI). To solve the problem, we proposed the optimized MTI with one wave plate, which can obtain cosine and sine functions by the combination of one wave plate and one linear polarizer. In the extraction of phase term using the combination of polarization components, the phase error occurs, and we analyzed such potential phase errors in the optimized MTI.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.5
• /
• pp.951-958
• /
• 2010

In this paper, we obtain the restricted maximum likelihood estimators (RMLE's) for means in normal distributions with the ordered mean constraints. The biases and mean squared errors (MSE's) of these RMLE's are approximated by Mote Carlo methods. In every case a substantial savings in MSE is obtained at the expense of a small loss in bias when using RMLE's instead of the unrestricted MLE's.

• Journal of the Optical Society of Korea
• /
• v.11 no.1
• /
• pp.10-17
• /
• 2007

We need two operation modes to obtain the complex hologram without bias and the conjugate image in the modified triangular interferometer (MTI). To solve the problem, we proposed the optimized MTI with one wave plate, which can obtain cosine and sine functions by the combination of one wave plate and one linear polarizer. In the extraction of phase term using the combination of polarization components, the phase error occurs, and we analyzed such potential phase errors in the optimized MTI.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.10 no.6
• /
• pp.234-240
• /
• 2002

This paper treats the problem of estimating the longitudinal velocity of a braking vehicle using measurements from an accelerometer and wheel speed data from standard anti-lock braking wheel speed sensors. We develop and experimentally test three velocity estimation algorithms of increasing complexity. The algorithm that works the best gives peak errors of less than 3 percent even when the accelerometer signal is significantly biased.

• Journal of the Korean Data and Information Science Society
• /
• v.15 no.1
• /
• pp.193-200
• /
• 2004

Parametric estimation of truncated point in a truncated exponential distribution will be considered. The MLE, bias reducing estimator and the ordinary jackknife estimator of the truncated parameter will be compared by mean square errors. And the MME and MLE of mean parameter and estimations of the right tail probability in the distribution will be compared by their MSE's.

• Journal of Institute of Control, Robotics and Systems
• /
• v.18 no.7
• /
• pp.632-637
• /
• 2012

This paper introduces a simple and efficient calibration method for three-axis accelerometers and three-axis gyroscopes using three-axis motion rate table. Usually, the performance of low cost MEMS-based inertial sensors is affected by scale and bias errors significantly. The calibration of these errors is a bothersome problem, but the previous calibration methods cannot propose simple and efficient method to calibrate the errors of three-axis inertial sensors. This paper introduces a new simple and efficient method for the calibration of accelerometer and gyroscope. By using a three-axis motion rate table, this method can calibrate the accelerometer and gyroscope simultaneously and simply. Experimental results confirm the performance of the proposed method.

• Journal of Korean Society for Atmospheric Environment
• /
• v.22 no.3
• /
• pp.271-285
• /
• 2006

A two-dimensional photochemical transport model (2D PTM) is simulated to describe the transport and chemical reaction of ozone related to aerosols in the troposphere and stratosphere. The vertical profiles and total amounts of ozone, which are advected by both residual Eulerian circulation and the adiabatic circulation under certain circumstance, have been compared with the observation data such as ozonesondes, Brewer spectrometer, the Upper Atmosphere Research Satellite (UARS), and the Total Ozone Mapping Spectrophotometer (TOMS). As a result, we find that the observed distribution of ozone Is adequately reproduced in the model at middle and high latitude in the Northern Hemisphere as well as at Phang ($36^{\circ}\;02'N,\;129^{\circ}\;23'E$) in South Korea. In particular, the 2D PTM is well simulated in the ozone decrease due to the Pinatubo volcanic eruption in 1991. However, ozone mixing ratio are more underestimated than those of UARS and ozonesondes, because are very sensitive to the latitude of transport across the tropopause associated with both Rummukainen errors and off-line model. Relative mean bias errors and relative root mean square errors of ozone calculations using the 2D PTM are shown within${\pm}10%$, respectively.