• 한국인공지능학회지
• /
• 제5권1호
• /
• pp.29-37
• /
• 2017

The range of problems that can be handled by the activation of big data and the development of hardware has been rapidly expanded and machine learning such as deep learning has become a very versatile technology. In this paper, mnist data set is used as experimental data, and the Cross Entropy function is used as a loss model for evaluating the efficiency of machine learning, and the value of the loss function in the steepest descent method is We applied the Gradient Descent Optimize algorithm to minimize and updated weight and bias via backpropagation. In this way we analyze optimal reliability value corresponding to the number of exercises and optimal reliability value without overfitting. And comparing the overfitting time according to the number of data changes based on the number of training times, when the training frequency was 1110 times, we obtained the result of 92%, which is the optimal reliability value without overfitting.

• 한국산학기술학회논문지
• /
• 제16권1호
• /
• pp.97-105
• /
• 2015

다중반응표면 최적화는 다수의 반응변수(품질특성치)를 최적화하는 입력변수의 조건을 찾는 것을 목적으로 한다. 다중반응표면 최적화를 위해 제안된 가중평균제곱오차(Weighted Mean Squared Error, WMSE) 최소화법은 평균제곱오차의 구성요소인 제곱편차와 분산에 서로 다른 가중치를 부여하는 방법이다. 지금까지 WMSE 최소화법과 관련하여, 개별 반응변수의 WMSE를 구성한 후 이들의 가중합을 최소화하는 가중합 기반 WMSE 최소화법이 제안되었다. 그러나 가중합 기반법은 목적함수 공간에서 볼록하지 않은 구간이 있고 이 구간에서 가장 선호되는 해가 존재할 경우 이 해를 찾아내지 못한다는 한계를 지니고 있다. 본 논문에서는 기존의 가중합 기반법의 한계점을 극복하기 위하여 Tchebycheff Metric 기반 WMSE 최소화법을 제안하고자 한다.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2008년도 학술발표회 논문집
• /
• pp.345-351
• /
• 2008

미래의 기후변화 영향평가에 있어 전지구모형(General Circulation Model)은 가장 중요한 자료 중 하나이다. 즉, 온실가스 방출(emission) 시나리오에 기초한 전지구모형의 모의결과를 이용하면 미래 수자원에 대한 정보를 얻을 수 있다. 하지만 미래 수자원은 방출 시나리오, 상세화(downscaling) 기법, 강우-유출모형, 전지구모형의 종류에 따라 크게 달라질 수 있어 매우 큰 불확실성(uncertainty)을 포함하고 있다. 이러한 불확실성을 줄이는 방법 중 하나로 전지구모형의 모의능력에 따라 가중치(weight)를 부여하고 결합(combining)하는 multi-model 앙상블(ensemble) 기법이 선진국을 중심으로 활발히 연구되고 있다. 본 연구에서는 우선 기후변화 영향평가를 위하여 국내에서 사용가능한 전지구모형을 조사하고 그 중CCSM3, CSRIO, ECHAM4, GFDL, MIRCO를 선택하였다. 한강 충주댐 유역에 대하여 과거($1980{\sim}1999$년)와 미래($2030{\sim}2049$년) 기간에 대하여 전지구모형의 기후정보를 간단한 선형보간법을 이용하여 상세화하였다. 다음으로 multi-model 앙상블 기법을 조사하였다. 본 연구에서는 Giorgi et al.(2002)이 제안한 Reliability Ensemble Average(REA) 기법을 적용하여 선형보간법으로 상세화한 전지구모형의 모의결과에 가중치를 주어 불확실성을 줄이는 연구를 수행하였다. 특히 REA를 구성하는 식 중 모형의 편차(bias) 뿐만 아니라 분산(variance)까지 고려함으로서 이를 개선하는 Modified-REA를 제안하였다. 제안한 방안을 이용하여 결합한 전지구모형의 모의결과가 기존 REA의 결과보다 기후정보의 불확실성을 더 줄일 수 있는 것으로 나타났다.

• International Journal of Advanced Culture Technology
• /
• 제5권2호
• /
• pp.82-89
• /
• 2017

The range of problems that can be handled by the activation of big data and the development of hardware has been rapidly expanded and machine learning such as deep learning has become a very versatile technology. In this paper, mnist data set is used as experimental data, and the Cross Entropy function is used as a loss model for evaluating the efficiency of machine learning, and the value of the loss function in the steepest descent method is We applied the GradientDescentOptimize algorithm to minimize and updated weight and bias via backpropagation. In this way we analyze optimal reliability value corresponding to the number of exercises and optimal reliability value without overfitting. And comparing the overfitting time according to the number of data changes based on the number of training times, when the training frequency was 1110 times, we obtained the result of 92%, which is the optimal reliability value without overfitting.

• 한국소음진동공학회논문집
• /
• 제19권11호
• /
• pp.1184-1191
• /
• 2009

A bearingless linear motor(BLLM) which consists of two stators and a common mover is able to levitate and move its mover without any linear bearing or even additive windings. In the previous study, BLLM was actively controlled on the translation and pitch motion, while the roll motion is passively stable. In order to control the roll motion, this paper suggests adding active magnetic bearings(AMBs) at bottom of the mover in BLLM. The AMBs control the roll motion and also partially supports the weight of the mover. In this paper, magnetic forces generated by the AMBs are estimated by using an FEM model. Based on the analysis results, the bias current of the AMBs is determined and a PD controller is designed. Through an experimental levitation test, it was verified that roll motion is well controlled by AMB during levitation.

• 대기
• /
• 제22권3호
• /
• pp.287-298
• /
• 2012

The predictability of heavy precipitation over the Korean Peninsula is studied using THORPEX Interactive Grand Global Ensemble (TIGGE) data. The performance of the six ensemble models is compared through the inconsistency (or jumpiness) and Root Mean Square Error (RMSE) for MSLP, T850 and H500. Grand Ensemble (GE) of the three best ensemble models (ECMWF, UKMO and CMA) with equal weight and without bias correction is consisted. The jumpiness calculated in this study indicates that the GE is more consistent than each single ensemble model. Brier Score (BS) of precipitation also shows that the GE outperforms. The GE is used for a case study of a heavy rainfall event in Korean Peninsula on 9 July 2009. The probability forecast of precipitation using 90 members of the GE and the percentage of 90 members exceeding 90 percentile in climatological Probability Density Function (PDF) of observed precipitation are calculated. As the GE is excellent in possibility of potential detection of heavy rainfall, GE is more skillful than the single ensemble model and can lead to a heavy rainfall warning in medium-range. If the performance of each single ensemble model is also improved, GE can provide better performance.

• Journal of Biosystems Engineering
• /
• 제23권6호
• /
• pp.561-570
• /
• 1998

The MLR(Multiple Linear Regression) models to estimate soluble solids content non-destructively were presented to make a selection of optimal photosensor utilized to measure the soluble solids content of apples. Visible and NIR absorbance in the 400 to 2498 nanometer(nm) wavelength region, soluble solids content(sugar content), hardness, and weight were measured for 400 apples(gala). Spectrophotometer with fiber optic probe was utilized for spectrum measurement and digital refractometer was used for soluble solids content. Correlation between absorbance spectrum and soluble solids content was analyzed to pick out the optimal wavelengths and to develop corresponding prediction model by means of MLR. For the coefficient of determination($R^2$) to be over 0.92, the MLR models out of the original absorbance were built based on 7 wavelengths of 992, 904, 1096, 1032, 880, 824, 1048nm, and the ones of the second derivative absorbance based on 5 wavelengths of 784, 1056, 992, 808, 872nm. The best model of the second derivative absorbance spectrum had $R^2$=0.91, bias= -0.02bx, SEP=0.28bx for unknown samples.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권9호
• /
• pp.4588-4608
• /
• 2017

The measurement of information security levels is a very important but difficult task. So far, various measurement methods have studied the development of new indices. Note, however, that researches have focused on the problem of attaining a certain level but largely neglecting research focused on the issue of how different types of possible flaws in security controls affect each other and which flaws are more critical because of these effects. Furthermore, applying the same weight across the board to these flaws has made it difficult to identify the relative importance. In this paper, the interrelationships among security flaws that occurred in the security controls of K-ISMS were analyzed, and the relative impact of each security control was measured. Additionally, a case-control study was applied using empirical data to eliminate subjective bias as a shortcoming of expert surveys and comparative studies. The security controls were divided into 2 groups depending on whether or not a security flaw occurs. The experimental results show the impact relationship and the severity among security flaws. We expect these results to be applied as good reference indices when making decisions on the removal of security flaws in an enterprise.

• 해양환경안전학회지
• /
• 제17권2호
• /
• pp.155-160
• /
• 2011

현재 지구온난화 등의 환경문제로 인해 각종 산업분야에서 정량화에 대한 요구가 증대되어 해양산업에도 그 수요가 증가하고 있는 실정이다. 따라서 본 연구에서는 차세대 경량화 재료인 마그네슘이 활용되기 위해서 반드시 극복해야할 가장 중요한 특성인 내식특성에 대하여 고찰하고, 그 내식특성 향상을 위한 마그네슘 박막의 Morphology나 결정배향성의 영향을 해명하고자 하였다. 실험결과로부터 제작한 Mg 박막의 전기화학적 내식특성은 Ar 가스압이 높은 조건에서 제작한 막일수록 내식특성이 우수하였다. 이러한 경향은 표면 및 단면의 Morphology와 결정배향성과의 상관관계를 통하여 설명 가능하였다.

• 한방비만학회지
• /
• 제4권1호
• /
• pp.213-219
• /
• 2004

Obesity is the condition that excess fat stored in the body. Nowadays, the prevalence of obesity has greatly increased due to westernized dietary habits and changes in lifestyle. It has been proved that obesity is a risk factor of many diseases. Many methods were developed to find curative means for obesity and exercise and controlling diet were considered very effective ones. In order to find the theraputic effect of the exercise and food diary, we examined patients treated by electric accupuncture and ear accupuncture because of obesity. We observed 16 patients with obesity who visited KyungHee Univ., Kangnam Hospital of Oriental Medicine from November, 2002 to March, 2003. To rule out an bias, we excluded the patients who were treated for other disease or took medicine. Exercise in hospital was not effective for treatment of obesity, but keeping diary of diet was effcetive for the decrease of body weight(p<0.05).