• The Journal of the Acoustical Society of Korea
• /
• v.25 no.4
• /
• pp.159-163
• /
• 2006

In this paper a data-driven method to improve the performance of the Jacobian adaptation (JA) for the noisy speech recognition is proposed. In stead of constructing the reference HMM by using the model composition method like the parallel model combination (PMC), we propose to train the reference HMM directly with the noisy speech. This was motivated from the idea that the directly trained reference HMM will model the acoustical variations due to the noise better than the composite HMM. For the estimation of the Jacobian matrices, the Baum-Welch algorithm is employed during the training. The recognition experiments have been done to show the improved performance of the proposed method over the Jacobian adaptation as well as other model compensation methods.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.1
• /
• pp.178-186
• /
• 1996

We develop a simple adaptive population coding model for neural networks based upon an error minimization method. Our model exhibits properties that have been experimentally observed in the population coding of the motor-cortical cells during the voluntary arm movements of primates. By removing a group of directionally tuned cells after learning, we study its contribution to the population coding. Through the learning process of the remained cells, we observe how the cells modify their preferred directions to reduce the coding errors. Since this adaptive property has been neither predicted nor experimentally observed before, it would be interesting to find whether a similar adaptive property exists in real cortices that are believed to encode the information in their cell populations.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.35 no.11
• /
• pp.990-998
• /
• 2007

Adaptive output feedback control technique using Neural Networks(NN) is proposed for uncertain nonlinear Multi-Input Multi-Output(MIMO) systems. Modified Dynamic Inversion Model(MDIM) is introduced to decouple uncertain nonlinearities from inversion-based control input. MDIM consists of approximated dynamic inversion model and inversion model error. One NN is applied to compensate the MDIM of the system. The output of the NN augments the tracking controller which is based upon a filtered error approximation with online weight adaptation laws which are derived from Lyapunov's direct method to guarantee tracking performance and ultimate boundedness. Several numerical results are illustrated in the simulation of Van der Pol system and unmanned helicopter with model uncertainties.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.6C
• /
• pp.893-900
• /
• 2004

While security traditionally has been an important issue in information systems, the problem of the greatest concern today is related to the availability of information and continuity of services. Since people and organizations now rely on distributed systems in accessing and processing critical services and mission, the availability of information and continuity of services are becoming more important. Therefore the importance of implementing systems that continue to function in the presence of security breaches cannot be overemphasized. One of the solutions to provide the availability and continuity of information system applications is introducing an intrusion tolerance system. Security mechanism and adaptation mechanism can ensure intrusion tolerance by protecting the application from accidental or malicious changes to the system and by adapting the application to the changing conditions. In this paper we propose an intrusion tolerance model that improves the developmental structure while assuring security level. We also design and implement an adaptive intrusion tolerance system to verify the efficiency of our model by integrating proper functions extracted from CORBA security modules.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.41 no.9
• /
• pp.817-823
• /
• 2017

In the past, research has been conducted on the development of an adaptive cruise control algorithm considering fuel efficiency, and an adaptive cruise control system considering fuel efficiency have been developed. However, research on optimizing vehicle and adaptive cruise control parameters in order to maximize performances is insufficient. In this study, the design optimization of vehicle and control parameters considering fuel efficiency, trackability, ride comfort and safe distance is performed. This paper proposes performance measures of vehicle behavior and develops an adaptive cruise control system. In addition, based on the screening of vehicle parameters that significantly influence performances, kriging surrogate models are constructed through a sequential design of experiment, and kriging surrogate model-based design optimization is performed to maximize fuel efficiency and satisfy target performances.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2003.05b
• /
• pp.1137-1140
• /
• 2003

네트워크의 QoS를 지원하기 위해서는 자원 관리에 적응제어구조의 도입이 필요하다. 이를 위해서는 사전에 네트워크의 대역폭 할당에 따른 전송률 응답특성을 구현하는 모델의 개발이 필수적이며 이 모델을 통하여 적응제어구조의 최적화를 진행해야 한다. 본 연구에서는 두 가지 방식의 모델을 제안하였다. 첫째는 동적 시스템 모델이며 다른 하나는 통계적 모델이다. 동적 시스템 모델은 네트워크의 동적 특성을 고려하여 도입하였으며, 통계적 모델은 측정된 전송률 데이터의 분포를 고려하여 도입하였다. 제시된 두 모델의 인자 결정을 위해 최적화 기법을 사용하였으며, 결과적으로 제시된 두 모델이 실제 네트워크의 동작과 유사함을 살펴보았다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10b
• /
• pp.130-132
• /
• 2001

신경망을 이용한 추천 기술은 항목이나 사용자간의 가중치를 학습할 수 있고, 자료 유형에 상관없이 데이터 처리가 용이하다. 또한 최근 연구를 통해서 그 우수성이 입증되고 있다. 그러나 사용자간의 상관관계로 추천하는 사용자 신경망 모델과 항목간의 상관관계로 추천하는 항목 신경망 모델이 서로 다른 관점으로 다른 선호도를 제시한 경우에 선택한 모델의 선호도에 따라 시스템의 성능이 좌우된다. 그러므로 효율적이고 성능이 우수한 추천 시스템을 위해 사용자와 항목 신경망 모델의 통합 방법을 제안한다. 두 모델 사이에 우선 순위를 결정하여 통합하는 순차적 통합 방법과 두 모델을 동시에 고려하는 병렬적 통합방법을 제안한다. 그러나 두 통합 방법은 선호도 예측 기준에 있어서 정적이고, 문제에 대한 적응성이 없다. 그러므로 신경망(퍼셉트론, 다층 퍼셉트론)을 이용한 통합 방법을 제안한다. 또한 퍼지의 소속함수를 이용하여 퍼지 추론를 적용한 통합 방법을 제안하고, 패턴 인식 분야에서 사용하는 BKS 방법을 적응하여 두 신경망 모델을 통합하여 실험한다. 본 논문에서는 사용자와 항목 신경망 모델을 통합함으로써 기존의 추천 기술인 연관 규칙과 단일 신경망 모델을 이용한 추천보다 우수함을 보이고 있다.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.6
• /
• pp.452-456
• /
• 2003

We discuss how to reduce the number of inverse matrix and its dimensions requested in MLLR framework for speaker adaptation. To find a smaller set of variables with less redundancy, we adapt PCA (principal component analysis) and ICA (independent component analysis) that would give as good a representation as possible. The amount of additional computation when PCA or ICA is applied is as small as it can be disregarded. 10 components for ICA and 12 components for PCA represent similar performance with 36 components for ordinary MLLR framework. If dimension of SI model parameter is n, the amount of computation of inverse matrix in MLLR is proportioned to O(n⁴). So, compared with ordinary MLLR, the amount of total computation requested in speaker adaptation is reduced by about 1/81 in MLLR with PCA and 1/167 in MLLR with ICA.

• Journal of the Korea Society of Computer and Information
• /
• v.12 no.1 s.45
• /
• pp.17-24
• /
• 2007

Critical practicality problems are cause to search the presentation and contents according to user request and purpose in previous internet system. Recently, there are a lot of researches about dynamic adaptable ontology based system. We designed ontology based educational system which uses discrete probability and user profile. This system provided advanced usability of contents by ontology and dynamic adaptive model based on discrete probability distribution function and user profile in ontology educational systems. This models represents application domain to weighted direction graph of dynamic adaptive objects and modeling user actions using dynamically approach method structured on discrete probability function. Proposed probability analysis can use that presenting potential attribute to user actions that are tracing search actions of user in ontology structure. This approach methods can allocate dynamically appropriate profiles to user.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.12
• /
• pp.3559-3568
• /
• 1999

The information overload with many information resources is an inevitable problem in modern electronic life. It is more difficult to search some information with user's information needs from an uncontrolled flood of many digital information resources, such as the internet which has been rapidly increased. So, many information retrieval systems have been researched and appeared. In text retrieval systems, they have met with user's information needs. While, in image retrieval systems, they have not properly dealt with user's information needs. In this paper, for resolving this problem, we proposed the intelligent user interface for image retrieval. It is based on HCOS(Human-Computer Symmetry) model which is a layed interaction model between a human and computer. Its' methodology is employed to reduce user's information overhead and semantic gap between user and systems. It is implemented with machine learning algorithms, decision tree and backpropagation neural network, for user adaptation capabilities of intelligent image retrieval system(IIRS).