• Annual Conference on Human and Language Technology
• /
• 1994.11a
• /
• pp.313-317
• /
• 1994

본 연구에서는 인간과 컴퓨터 사이의 음성을 이용한 대화 시스템을 구현하였다. 특별히 음성을 인식하는데 있어서 단어추출(word apotting) 방법을 사용하는 경우에 알맞은 의미 분석 방법과 도표 형태의 규칙을 기반으로 하여 시스템의 응답을 생성하는 방법에 대하여 연구하였다. 단어추출 방법을 사용하여 음성을 인식하는 경우에는 형태소분석 및 구문분석의 과정을 이용하여 사용자의 발화 의도를 분석하기 어려우므로 새로운 의미분석 방법을 필요로 한다. 본 연구에서는 퍼지 관계를 사용하여 사용자의 발화 의도를 파악하는 새로운 의미분석 방법을 제안하였다. 그리고, 사용자의 발화 의도에 적절한 시스템의 응답을 만들고 응답의 내용을 효율적으로 관리하기 위한 방범으로 현재의 상태와 사용자의 의도에 따른 응답 규칙을 만들었다. 이 규칙은 도표의 형태로 구현되어 규칙의 갱신 및 확장을 편리하게 만들었다. 대화의 영역은 열차 예매에 관련된 예매, 취소, 문의 및 관광지 안내로 제안하였다. 음성의 오인식에 의한 오류에 적절히 대처하기 위해 시스템의 응답은 확인 및 수정 과정을 포함하고 있다. 본 시스템은 문자 입력과 음성 입력으로 각각 실험한 결과, 사용자는 시스템의 도움을 받아 자신이 의도하는 목적을 달성할 수 있었다.

• Proceedings of the KIEE Conference
• /
• 2003.07d
• /
• pp.2750-2752
• /
• 2003

흑점은 태양 표면에 검은 구멍처럼 보이는 것으로 흑점이 나타나면 태양활동이 활발함을 의미한다. 이러한 태양활동은 플레어나 홍염 등의 형태로 표출되어 지구의 자기장을 변동시킴으로써 전력, 통신 시스템의 장애를 유발하게 된다. 따라서 이러한 흑점 데이터를 예측함으로써 사전에 대비할 수 있도록 할 필요가 있다. 흑점 시계열 데이터의 예측에 사용된 시스템은 병렬구조를 갖는 퍼지시스템(PSFS)으로 각 퍼지시스템의 규칙은 주어진 입출력 데이터를 클러스터링하여 생성하였다. 특히, 흑점 시계열 데이터와 같이 주기성향을 갖는 테이터의 경우에도 적용가능하도륵 유연한 구조를 갖는 개선된 PSFS를 제안하여 그 성능을 검증하였다.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.11 no.1
• /
• pp.125-132
• /
• 2011

Fire monitoring system detects a fire based on the values of various sensors, such as smoke, CO, temperature, or change of temperature. It detects a fire by comparing sensed values with predefined threshold values for each sensor. However, to prevent a fire it is required to predict a situation which has a possibility of fire occurrence. In this work, we propose a fire recognition system using a fuzzy inference method. The rule base is constructed as a combination of fuzzy variables derived from various sensed values. In addition, in order to solve generalization and formalization problems of rule base construction from expert knowledge, we analyze features of fire patterns. The constructed rule base results in an improvement of the recognition accuracy. A fire possibility is predicted as one of 3 levels(normal, caution, danger). The training data of each level is converted to fuzzy rules by FCM(fuzzy C-means clustering) and those rules are used in the inference engine. The performance of the proposed approach is evaluated by using forest fire data from the UCI repository.

• Journal of Internet Computing and Services
• /
• v.12 no.4
• /
• pp.107-117
• /
• 2011

The appearance of Web has brought an substantial revolution to all fields of society such knowledge management and business transaction as well as traditional information retrieval. In this paper, we propose an EFASIT(Extended Fuzzy AHP and SImilarity Technology) model considering the emotion analysis. And we combine the Extended Fuzzy AHP Method(EFAM) with SImilarity Technology(SIT) based on the domain corpus information in order to efficiently retrieve the document on the Web. The proposed the EFASIT model can generate the more definite rule according to integration of fuzzy knowledge of various decision-maker, and can give a help to decision-making, and confirms through the experiment.

• Journal of Internet Computing and Services
• /
• v.10 no.5
• /
• pp.127-133
• /
• 2009

This paper presents an approach to classify normal and epilepsy from electroencephalogram(EEG) using a neural network with weighted fuzzy membership functions(NEWFM). To extract input features used in NEWFM, wavelet transform is used in the first step. In the second step, the frequency distribution of signal and the amount of changes in frequency distribution are used for extracting twenty-four numbers of input features from coefficients and approximations produced by wavelet transform in the previous step. NEWFM classifies normal and epilepsy using twenty four numbers of input features, and then the accuracy rate is 98%.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2018.10a
• /
• pp.771-773
• /
• 2018

인공지능(AI), 사물인터넷(IoT)등의 4차 산업기술은 철도안전의 핵심수단으로 부상하고 있으며 차량, 위험관리, 운행관리, 보안관리 등의 점진적인 적용분야 확장을 통해 철도안전에 대한 신뢰성을 향상시킬 수 있는 방안에 대한 관심이 집중되고 있다. 본 논문에서는 IoT 기반의 다양한 철도인프라 데이터를 활용하여 열차주행상태에 영향을 줄 수 있는 이상상황 인식 모델 및 열차자율주행을 위한 제어기술에 필요한 정보로 인프라 상태를 제공하는 방식을 제안한다. 철도 인프라 상황인지에 필요한 데이터는 레일온도, 선로 지정물, 승객 수, 선로 적설량을 지정하였고, 제안 인식모델의 스게노 퍼지추론 방식을 적용한 후 철도차량 운전관련 취급규정 및 취급세척을 기반으로 퍼지규칙(Fuzzy Rule)을 15개 생성하였다. 인프라데이터셋을 활용하여 제안모델의 인식률 평가에 사용하였으며 인식률 결과는 약 86%의 정확성을 보였다. 퍼지추론 기반 방식의 철도인프라 이상상태 인식모델을 철도분야에 접목시킨다면 기존의 관제기반 방식보다 효율적인 철도인프라 상황인식이 가능할 것으로 판단된다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.5
• /
• pp.533-538
• /
• 2004

This paper proposes a systematic method to develop short-term electrical load forecasting systems using neuro-fuzzy models. The proposed system predicts the electrical loads with the lead times of 1 hour, 24 hour, and 168 hour. To do so, the load forecasting system first builds an initial structure off-line for each hour of four day types and then stores the resultant initial structures in the initial structure bank. 96 initial structures are constructed for each prediction lead time. Whenever a prediction needs to be made, the proposed system initializes the neuro-fuzzy model with the appropriate initial structure stored and trains the initialized prediction modell. To improve the performance of the prediction system in terms of accuracy and reliability at the same time, the prediction model employs only two inputs. It makes possible to interpret the fuzzy rules to be learned. In order to demonstrate the viability of the proposed method, we develop a load forecasting system by using the real load data collected during 1996 and 1997 at KEPCO. Simulation results reveal that the prediction system developed in this paper can achieve a remarkable improvement on both accuracy and reliability

• Journal of the Korean Institute of Intelligent Systems
• /
• v.5 no.4
• /
• pp.41-55
• /
• 1995

In this paper, an optimal idenfication method using fuzzy-neural networks is proposed for modeling of nonlinear complex systems. The proposed fuzzy-neural modeling implements system structure and parameter identification using the intelligent schemes together wlth optimization theory, linguistic fuzzy implication rules, and neural networks(NNs) from input and output data of processes. Inference type for this fuzzy-neural modeling is presented as simplified inference. To obtain optimal model, the learning rates and momentum coefficients of fuzzy-neural networks(FNNs) are tuned automatically using improved modified complex method and modified learning algorithm. For the purpose of its application to nonlinear processes, data for route choice of traffic problems and those for activateti sluge process of sewage treatment system are used for the purpose of evaluating the performance of the proposed fuzzy-neural network modeling. The results show that the proposed method can produce the intelligence model with higher accuracy than other works achieved previously.

• Journal of KIISE:Software and Applications
• /
• v.31 no.9
• /
• pp.1171-1179
• /
• 2004

We describe word boundary detection that extracts the boundary between speech and non-speech. The proposed method uses two features. One is the normalized root mean square of speech signal, which is insensitive to white noises and represents temporal information. The other is the normalized met-frequency band energy of voice signal, which is frequency information of the signal. Our method detects word boundaries using a recurrent fuzzy associative memory(RFAM) that extends FAM by adding recurrent nodes. Hebbian learning method is employed to establish the degree of association between an input and output. An error back-propagation algorithm is used for teaming the weights between the consequent layer and the recurrent layer. To confirm the effectiveness, we applied the suggested system to voice data obtained from KAIST.

• Journal of KIISE:Software and Applications
• /
• v.33 no.8
• /
• pp.692-702
• /
• 2006

In this study, a weighted fuzzy min-max (WFMM) neural network model for pattern classification is proposed. The model has a modified structure of FMM neural network in which the weight concept is added to represent the frequency factor of feature values in a learning data set. First we present in this paper a new activation function of the network which is defined as a hyperbox membership function. Then we introduce a new learning algorithm for the model that consists of three kinds of processes: hyperbox creation/expansion, hyperbox overlap test, and hyperbox contraction. A weight adaptation rule considering the frequency factors is defined for the learning process. Finally we describe a feature analysis technique using the proposed model. Four kinds of relevance factors among feature values, feature types, hyperboxes and patterns classes are proposed to analyze relative importance of each feature in a given problem. Two types of practical applications, Fisher's Iris data and Cleveland medical data, have been used for the experiments. Through the experimental results, the effectiveness of the proposed method is discussed.