• 지능정보연구
• /
• 제7권2호
• /
• pp.51-63
• /
• 2001

데이터로부터 학습하여 룰을 추출하는 귀납적 학습기법은 데이터 마이닝의 주요 도구 중 하나이다. 귀납적 학습 기법은 불필요한 변수나 잡음이 섞인 변수를 포함하여 학습하는 경우 생성된 룰의 예측 성능이 떨어지고 불필요하게 룰이 복잡하게 구성될 수 있다. 따라서 귀납적 학습 기법의 예측력을 높이고 룰의 구성도 간단하게 할 수 있는 주요 변수 부분집합을 선정하는 방안이 필요하다. 귀납적 학습에서 예측력을 높이기 위해 많이 사용되는 부분집합 선정을 위한 포장 기법은 최적의 부분집합을 찾기 위해 전체 부분집합을 탐색한다. 이때 전체 변수의 수가 많아지면 부분집합의 탐색 공간이 너무 커져서 탐색하기 어려운 문제가 된다. 본 연구에서는 포장 기법에 신경망 민감도 분석을 결합한 귀납적 학습 기법의 변수 부분집합 선정 방안을 제시한다. 먼저, 신경망의 민감도 분석 기법을 이용하여 전체 변수를 중요도 순으로 순서화 한다. 다음에 순서화된 정보를 이용하여 귀납적 학습 기법의 예측력을 높일 수 있는 부분집합을 찾아 나간다. 제안된 방법을 세 데이터 셋에 적용한 결과 일정한 반복 회수 이내에 예측력이 향상된 부분집합을 얻을 수 있음을 볼 수 있다.

• 디지털융복합연구
• /
• 제11권10호
• /
• pp.47-57
• /
• 2013

빅데이터 분석은 데이터의 양, 처리속도, 다양성 측면에서 데이터 마이닝과 달리 문제해결과 의사결정을 위해서는 새로운 도구를 필요로 한다. 많은 글로벌 IT기업들은 사용하기 쉽고 기능성이 우수한 모델링 능력을 가진 다양한 빅데이터 제품을 출시하고 있다. 빅데이터 패키지는 분석도구, 인프라, 플랫폼 형태로 하드웨어와 소프트웨어를 포함한 솔루션이다. 빅데이터의 수집, 저장, 분석, 시각화가 가능한 제품이다. 빅데이터 패키지는 업체별로 제품 종류가 많고 복잡한 기능을 가질 뿐만 아니라 선정에 있어서 전문 지식을 필요로 하며 일반적인 소프트웨어 패키지보다 그 중요성이 높기 때문에 의사결정 방법의 개발이 요구된다. 본 연구는 빅데이터 패키지 도입을 위한 의사결정지원방법을 제안하는 것이 목표이다. 문헌적 고찰을 통하여 빅데이터 패키지의 특징과 기능을 비교하고, 선정기준을 제안한다. 패키지 도입 타당성을 평가하기 위하여 비용과 혜택 각각을 목표노드로 하는 AHP 모델 및 선정기준을 목표노드로 하는 AHP 모델을 제안하고 이들을 결합하여 최적의 패키지를 선정하는 과정을 보인다.

• Journal of Electrical Engineering and Technology
• /
• 제13권6호
• /
• pp.2145-2157
• /
• 2018

Congestion Management (CM) is an attractive research area in the electrical power transmission with the power compensation abilities. Reconfiguration and the Flexible Alternating Current Transmission Systems (FACTS) devices utilization relieve the congestion in transmission lines. The lack of optimal power (real and reactive) usage with the better transfer capability and minimum cost is still challenging issue in the CM. The prediction of suitable place for the energy resources to control the power flow is the major requirement for power handling scenario. This paper proposes the novel optimization principle to select the best location for the energy resources to achieve the real-reactive power compensation. The parameters estimation and the selection of values with the best fitness through the Symmetrical Distance Travelling Optimization (SDTO) algorithm establishes the proper controlling of optimal power flow in the transmission lines. The modified fitness function formulation based on the bus parameters, index estimation correspond to the optimal reactive power usage enhances the power transfer capability with the minimum cost. The comparative analysis between the proposed method with the existing power management techniques regarding the parameters of power loss, cost value, load power and energy loss confirms the effectiveness of proposed work in the distributed renewable energy systems.

• 한국소음진동공학회논문집
• /
• 제14권12호
• /
• pp.1330-1337
• /
• 2004

In this study, the equations of motion of a rigid rotor supported by magnetic bearings are derived via Hamilton's principle, and transformed to a state-space form for control purpose. The optimal motion control of rotor magnetic bearing system based on the LQR(linear quadratic regulator) theory is addressed. New schemes related to the selection of the state weighting matrix Q and the control weighting matrix R involved in the quadratic functional to be minimized are proposed. And the robust control of the system with an LMI(linear matrix inequality) based H$_{\infty}$ theory is dealt with in this paper. Loop shapings of TFM (transfer function matrix) are used to increase the performance of control capability of the system. The control abilities of LQR and H$_{\infty}$ controller are compared by simulation and experimental tests and show that the capability of H$_{\infty}$ controller is superior to that of LQR.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2005년도 ICCAS
• /
• pp.1505-1509
• /
• 2005

This paper introduces a new architecture of genetically optimized self-organizing fuzzy polynomial neural networks by means of information granulation. The conventional SOFPNNs developed so far are based on mechanisms of self-organization and evolutionary optimization. The augmented genetically optimized SOFPNN using Information Granulation (namely IG_gSOFPNN) results in a structurally and parametrically optimized model and comes with a higher level of flexibility in comparison to the one we encounter in the conventional FPNN. With the aid of the information granulation, we determine the initial location (apexes) of membership functions and initial values of polynomial function being used in the premised and consequence part of the fuzzy rules respectively. The GA-based design procedure being applied at each layer of genetically optimized self-organizing fuzzy polynomial neural networks leads to the selection of preferred nodes with specific local characteristics (such as the number of input variables, the order of the polynomial, a collection of the specific subset of input variables, and the number of membership function) available within the network. To evaluate the performance of the IG_gSOFPNN, the model is experimented with using gas furnace process data. A comparative analysis shows that the proposed IG_gSOFPNN is model with higher accuracy as well as more superb predictive capability than intelligent models presented previously.

• 한국조리학회지
• /
• 제24권2호
• /
• pp.23-33
• /
• 2018

Food adulteration and food fraud should not be neglected. The present study aimed to investigate the awareness of adulterated food and its management beliefs and capabilities among teenagers' parents. Data were collected from 425 adolescents' parents having different levels of income and education. The results of factor analysis indicated that adulterated food management beliefs was classified into attitude, necessity, and anxiety. The adulterated food management capability was sub-grouped into hygiene and nutrition, knowledge, citizen action and environmental grasp. The adulterated food management capabilities were significantly different according child's school, education level and monthly income (p<0.05). The attitude factor of adulterated food management beliefs appeared to have a significant (p<0.05) impact on all factors of adulterated food management capabilities, however the necessity factor had a significant (p<0.001) impact only on factor of hygiene and nutrition. The results of the present study suggested that parents need to be aware themselves as well as to teach their children about right food selection and consumption. The findings of the study might be useful in government policy planning regarding the public health issues and dietary education of adolescents and parents.

• Journal of Power Electronics
• /
• 제15권1호
• /
• pp.18-30
• /
• 2015

This paper focuses on a systematical and in-depth analysis of the reactive power and soft-switching regions of Dual Active Bridge (DAB) converters with dual-phase-shift (DPS) control to achieve high efficiency in a wide operating range. The key features of the DPS operating modes are characterized and verified by analytical calculation and experimental tests. The mathematical expressions of the reactive power are derived and the reductions of the reactive power are illustrated with respect to a wide range of output power and voltage conversion ratios. The ZVS soft-switching boundary of the DPS is presented and one more leg with ZVS capability is achieved compared with the CPS control. With the selection of the optimal operating mode, the optimal phase-shift pair is determined by performance indices, which include the minimum peak or rms inductor current. All of the theoretical analysis and optimizations are verified by experimental tests. The experimental results with the DPS demonstrate the efficiency improvement for different load conditions and voltage conversion ratios.

• Journal of Information Processing Systems
• /
• 제15권4호
• /
• pp.986-1016
• /
• 2019

Mobile user interface pattern (MUIP) is a kind of structured representation of interaction design knowledge. Several studies have suggested that MUIPs are a proven solution for recurring mobile interface design problems. To facilitate MUIP selection, an effective clustering method is required to discover hidden knowledge of pattern data set. In this paper, we employ the semi-supervised kernel fuzzy c-means clustering (SSKFCM) method to cluster MUIP data. In order to improve the performance of clustering, clustering parameters are optimized by utilizing the global optimization capability of particle swarm optimization (PSO) algorithm. Since the PSO algorithm is easily trapped in local optima, a novel PSO algorithm is presented in this paper. It combines an improved intuitionistic fuzzy entropy measure and a new population search strategy to enhance the population search capability and accelerate the convergence speed. Experimental results show the effectiveness and superiority of the proposed clustering method.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권11호
• /
• pp.5496-5521
• /
• 2018

Object detection and tracking is the basic capability of mobile robots to achieve natural human-robot interaction. In this paper, an object tracking system of mobile robot is designed and validated using improved multiple instance learning algorithm. The improved multiple instance learning algorithm which prevents model drift significantly. Secondly, in order to improve the capability of classifiers, an active sample selection strategy is proposed by optimizing a bag Fisher information function instead of the bag likelihood function, which dynamically chooses most discriminative samples for classifier training. Furthermore, we integrate the co-training criterion into algorithm to update the appearance model accurately and avoid error accumulation. Finally, we evaluate our system on challenging sequences and an indoor environment in a laboratory. And the experiment results demonstrate that the proposed methods can stably and robustly track moving object.

• 한국군사과학기술학회지
• /
• 제26권2호
• /
• pp.197-204
• /
• 2023

The importance of each warfighting function for Army TIGER unit attack drones is measured using the AHP technique. As a result, the importance of attack drones is high in the order of maneuver, firepower, intelligence, command/control, protection, and operation sustainment, but the importance of maneuver, firepower, and intelligence are almost similar. In addition, it is analyzed that attack drones capable of carrying out day and night missions by being equipped with an EO/IR sensor and being commanded/controlled in conjunction with the C4I system to eliminate threats with small bombs or aircraft collisions is needed. Finally, based on the results of this study, a virtual battle scenario for attack drones is proposed.