• Journal of the Korean Institute of Intelligent Systems
• /
• v.7 no.4
• /
• pp.82-92
• /
• 1997

In this paper, a fuzzy measure and integral-based model fnr interactive systems is proposed. The processes of model identification consists of the following three steps : (i) structure identification (ii) parameter identification and (iii) selection of an optimal model. An algorithm for the model structure identification using the well-known genetic algorithm ((;A) with a modified selection operator is proposed. A method for the identification of par;imetcrs corresponding to fuzzy measures is presented. A statistical model selection criterion is used for the selection of an optimal model among the candidates. Finally, experimental results obtained hy applying the proposed model to the subjective evaluation data set and the well-known time series data are presented to show the validity of the proposed model.

• Defense and Technology
• /
• no.8 s.282
• /
• pp.46-55
• /
• 2002

지난 7월호에서는 COCOMO 모델, 기능점수모델, PRICE-S모델, 일반 COSDES 모델 등의 무기체계 S/W 개발비용 산정모델들을 조사 및 분석하여 제시하였다. 이번 8월호에서는 무기체계 소프트웨어 개발비용산저에 영향을 미치는 요소를 기존의 '한소협' 모델, PRICE-S 모델, 기능점수 모델, COCOMO 모델 등 여러 소프트웨어 개발비용 산정모델을 참고 후 무기체계의 특성을 고려하여 식별하였다. 또한 식별된 요소들을 '한소협' 모델의 절차와 방법에 따라 무기체계 소프트웨어의 스텝 수 산정 영향요소, 환경요인 보정계수 영향요소, 제경비 및 기술료 산정 영향요소들을 식별하여 제시하였다.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.3
• /
• pp.125-138
• /
• 2016

The service identification in service-oriented developments has been conducted by based on workflow, goals, scenarios, usecases, components, features, and patterns. However, the identification of service by semantic approach at the business value view was not detailed yet. In order to enhance accuracy of identifying business service, this paper proposes a method for identifying business service by analyzing syntax and semantics in XL-BPMN model. The business processes based on business scenario are identified, and they are designed in a XL-BPMN business process model. In this business process model, an unit business service is identified through binding closely related activities by the integrated analysis result of syntax patterns and properties-based semantic similarities between activities. The method through XL-BPMN model at upper business levels can identify the reusable unit business service with high accuracy and modularity. It also can accelerate more service-oriented developments by reusing identified services.

• ICROS
• /
• v.18 no.3
• /
• pp.37-42
• /
• 2012

이번 편에서는 첫 번째 편에서 소개된 블랙박스 모델을 결정하는 기법 중의 하나인 매개변수 시스템 식별법을 소개한다. 이 기법은 식별하고자 하는 대상 시스템에 대하여 매개변수들로 표현되는 여러 가지 후보 모델들을 선정한 후 확정된 입출력 데이터와 추정 알고리즘을 적용하고 여러 가지 검증과정을 통하여 실제 시스템의 데이터에 가장 가까운 특성을 보이는 모델을 선정하는 방법이다. 이를 위해서 본 편에서는 선형-시불변 시스템의 블랙박스 식별에서 종종 사용되는 여러 가지 모델구조들을 소개한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2016.10a
• /
• pp.297-300
• /
• 2016

최근 빅데이터 산업이 발전하고 있는 상황에서 빅데이터 산업에 활용되는 개인정보의 보호에 관한 문제가 대두하고 있다. 빅데이터 산업에서 개인정보를 활용하기 위해서는 비식별화 조치를 해야 한다. 하지만 비식별화는 비식별화 평가 모델 자체의 취약성과 더불어 비식별화된 개인정보를 재식별화 하는 위험성도 존재한다. 본 논문은 적정성 평가 모델, 비식별화 조치 기술, 재식별에 관한 위험성을 연구하고 각 위험성에 대한 대응 방안을 통해 재식별화의 문제를 해결하여 빅데이터 산업에서 비식별화된 개인정보가 안전히 쓰일 수 있도록 해야 한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2017.11a
• /
• pp.285-288
• /
• 2017

비식별화 모델은 데이터 공유를 위한 모델로 원본데이터를 비식별화 변환 처리하여 개인정보를 보호함과 동시에 분석에 필요한 데이터를 외부에 제공하는 모델로 연구되어 왔다. 변환 방법으로는 삭제, 일반화, 범주화 기술 등이 주로 사용되며 변환 과정 중에는 재식별 가능성을 최소화하기 위해 k-익명성, l-다양성, t-근접성 혹은 differential privacy 등의 프라이버시 모델이 적용되고 있다. 하지만 변환된 비식별 데이터 세트는 필연적으로 원본 데이터 세트와 다른 값을 가지며 이는 결과적으로 최종 분석 결과에 영향을 주게 된다. 이를 위해 두 데이터 세트 간의 차이를 상이도(dissimilarity) 혹은 정보 손실율(information loss)이라는 지표로 측정 하고 있으며 본 지표는 비식별 데이터의 활용성을 평가 하는 데에 매우 중요한 역할을 한다. 본 연구에서는 비식별 데이터와 원본 데이터와 간의 차이를 도메인 기반의 절대적인 기준대비로 표현한 상이도 측정 방법을 제안하며, 그 유효성을 실데이터 기반의 실험을 통해 검증하였다.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.8
• /
• pp.735-743
• /
• 2002

In this paper, we propose three new text-independent speaker identification methods. At first, to exclude the frames not having enough features of speaker's vocal from calculation of the maximum likelihood, we propose the FS(Frame Selection) method. This approach selects the important frames by evaluating the difference between the biggest likelihood and the second in each frame, and uses only the frames in calculating the score of likelihood. Our secondly proposed, called the Hybrid, is a combined version of the FS and WMR(Weighting Model Rank). This method determines the claimed speaker using exponential function weights, instead of likelihood itself, only on the selected frames obtained from the FS method. The last proposed, called MWMR (Modified WMR), considers both original likelihood itself and its relative position, when the claimed speaker is determined. It is different from the WMR that take into account only the relative position of likelihood. Through the experiments of the speaker identification, we show that the all the proposed have higher identification rates than the ML. In addition, the Hybrid and MWMR have higher identification rate about 2% and about 3% than WMR, respectively.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.6
• /
• pp.457-462
• /
• 2003

In this paper, we propose a speaker pruning method for real-time processing and improving performance of speaker identification system based on GMM(Gaussian Mixture Model). Conventional speaker identification methods, such as ML (Maximum Likelihood), WMR(weighting Model Rank), and MWMR(Modified WMR) we that frame likelihoods are calculated using the whole frames of each input speech and all of the speaker models and then a speaker having the biggest accumulated likelihood is selected. However, in these methods, calculation cost and processing time become larger as the increase of the number of input frames and speakers. To solve this problem in the proposed method, only a part of speaker models that have higher likelihood are selected using only a part of input frames, and identified speaker is decided from evaluating the selected speaker models. In this method, fm can be applied for improving the identification performance in speaker identification even the number of speakers is changed. In several experiments, the proposed method showed a reduction of 65% on calculation cost and an increase of 2% on identification rate than conventional methods. These results means that the proposed method can be applied effectively for a real-time processing and for improvement of performance in speaker identification.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.3 no.4
• /
• pp.47-60
• /
• 1999

The stability and efficiency of structural control systems depend on the accuracy of mathematical model of the system to be controlled. In this study, state equation models of a small scale test structure and an AMD(active mass damper) are obtained separately using OKID(observer/Kalman filter identification) which is a time domain system identification method. The test structure with each floor acceleration as outputs is identified for two inputs - the ground acceleration and the acceleration of the moving mass of AMD relative to the installation floor - individually and the two identified state equation models are integrated into one by model reduction method. The AMD is identified with the motor control signal as an input and the relative acceleration of the moving mass as an output, and it is shown that the identified model has large damping ratio and phase shift. The transfer functions and the time histories reconstructed from the identified models of the test model and the AMD match well with those measured from the experiment.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.14 no.5
• /
• pp.416-423
• /
• 2004

In this paper, an experimental verification of system identification technique for constructing finite element model is conducted for a three-story test structure equipped with an active mass driver (AMD). Twenty Gaussian white noises were used as the input for AMD, and the corresponding accelerations of each floor are measured. Then, the complex frequency response function (FRF) for the input, the force induced by the AMD, was obtained and subsequently, the Markov parameters and system matrices were estimated. The magnitudes as well as phase of experimentally obtained FRFs match well with those of analytically obtained FRFs.