• 대한전기학회논문지
• /
• 제45권2호
• /
• pp.325-328
• /
• 1996

In this paper, we proposed an estimation method of a posterior probability and PDF(Probability density function) using a feed forward neural network and code books of VQ(vector quantization). In this study, We estimates a posterior probability and probability density function, which compose a new parameter with well-known Mel cepstrum and verificate the performance for the five vowels taking from syllables by NN(neural network) and PNN(probabilistic neural network). In case of new parameter, showed the best result by probabilistic neural network and recognition rates are average 83.02%.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2003년도 봄 학술발표회 논문집
• /
• pp.357-365
• /
• 2003

The major problems with the conventional neural network, especially Back Propagation Neural Network, arise from the necessity of many training data for neural network learning and ambiguity in the relation of neural network structure to the convergence of solution. In this paper, the PNN is used as a pattern classifier to detect the damage of structure to avoid those drawbacks of the conventional neural network. In the PNN-based pattern classification problems, the probability density function for patterns is usually assumed by Gaussian distribution. But, in this paper, several probability density functions are investigated in order to select the most approriate one for structural damage assessment.

• 통합자연과학논문집
• /
• 제13권4호
• /
• pp.147-151
• /
• 2020

In this paper, we propose a method to detect concept drift by applying Convolutional Neural Network (CNN) in a data stream environment. Since the conventional method compares only the final output value of the CNN and detects it as a concept drift if there is a difference, there is a problem in that the actual input value of the data stream reacts sensitively even if there is no significant difference and is incorrectly detected as a concept drift. Therefore, in this paper, in order to reduce such errors, not only the output value of CNN but also the probability vector are used. First, the data entered into the data stream is patterned to learn from the neural network model, and the difference between the output value and probability vector of the current data and the historical data of these learned neural network models is compared to detect the concept drift. The proposed method confirmed that only CNN output values could be used to reduce detection errors compared to how concept drift were detected.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2004년도 ICCAS
• /
• pp.631-634
• /
• 2004

In this paper, the Bayesian recurrent neural network (BRNN) is proposed to predict time series data. Among the various traditional prediction methodologies, a neural network method is considered to be more effective in case of non-linear and non-stationary time series data. A neural network predictor requests proper learning strategy to adjust the network weights, and one need to prepare for non-linear and non-stationary evolution of network weights. The Bayesian neural network in this paper estimates not the single set of weights but the probability distributions of weights. In other words, we sets the weight vector as a state vector of state space method, and estimates its probability distributions in accordance with the Bayesian inference. This approach makes it possible to obtain more exact estimation of the weights. Moreover, in the aspect of network architecture, it is known that the recurrent feedback structure is superior to the feedforward structure for the problem of time series prediction. Therefore, the recurrent network with Bayesian inference, what we call BRNN, is expected to show higher performance than the normal neural network. To verify the performance of the proposed method, the time series data are numerically generated and a neural network predictor is applied on it. As a result, BRNN is proved to show better prediction result than common feedforward Bayesian neural network.

• 한국구조물진단유지관리공학회 논문집
• /
• 제6권4호
• /
• pp.169-179
• /
• 2002

This paper presents an efficient algorithm for the estimation of damage location and severity in structure using Probabilistic Neural Network (PNN). Artificial neural network has been being used for damage assessment by many researchers, but there are still some barriers that must be overcome to improve its accuracy and efficiency. The major problems with the conventional neural network are the necessity of many training data for neural network learning and ambiguity in the relation of neural network architecture with convergence of solution. In this paper, PNN is used as a pattern classifier to overcome those problems in the conventional neural network. The basic idea of damage assessment algorithm proposed in this paper is that modal characteristics from a damaged structure are compared with the training patterns which represent the damage in specific element to determine how close it is to training patterns in terms of the probability from PNN. The training pattern that gives a maximum probability implies that the element used in producing the training pattern is considered as a damaged one. The proposed damage assessment algorithm using PNN is applied to a 2-span continuous beam model structure to verify the algorithm.

• 제어로봇시스템학회논문지
• /
• 제10권12호
• /
• pp.1295-1304
• /
• 2004

In this paper, the Bayesian recurrent neural network is proposed to predict time series data. A neural network predictor requests proper learning strategy to adjust the network weights, and one needs to prepare for non-linear and non-stationary evolution of network weights. The Bayesian neural network in this paper estimates not the single set of weights but the probability distributions of weights. In other words, the weights vector is set as a state vector of state space method, and its probability distributions are estimated in accordance with the particle filtering process. This approach makes it possible to obtain more exact estimation of the weights. In the aspect of network architecture, it is known that the recurrent feedback structure is superior to the feedforward structure for the problem of time series prediction. Therefore, the recurrent neural network with Bayesian inference, what we call Bayesian recurrent neural network (BRNN), is expected to show higher performance than the normal neural network. To verify the proposed method, the time series data are numerically generated and various kinds of neural network predictor are applied on it in order to be compared. As a result, feedback structure and Bayesian learning are better than feedforward structure and backpropagation learning, respectively. Consequently, it is verified that the Bayesian reccurent neural network shows better a prediction result than the common Bayesian neural network.

• Ocean and Polar Research
• /
• 제25권spc3호
• /
• pp.347-351
• /
• 2003

A new approach to reliability analysis of rubble mound breakwater using neural network is proposed. At first, a neural network model which can estimate the stability number of any breakwaters for some design conditions is trained. Then, the neural network model is integrated with Monte Carlo simulation technique in order to calculate probability of failure for the breakwater. The proposed technique is compared with conventional approach using empirical formula.

• 한국통신학회논문지
• /
• 제31권9C호
• /
• pp.853-858
• /
• 2006

음성인식에서 TMHMM(Tied Mixture Hidden Markov Model)은 자유 매개변수의 수를 감소시키기 위한 좋은 접근이지만, GPDF(Gaussian Probability Density Function) 군집화 오류에 의해 음성인식의 오류를 발생시켰다. 본 논문은 TMHMM에서 발생하는 군집화 오류를 최소화하기 위하여 HCNN(Homogeneous Centroid Neural Network) 군집화 알고리즘을 제안한다. 제안된 알고리즘은 CNN(Centroid Neural Network)을 TMHMM상의 음향 특징벡터에 활용하였으며, 다른 상태에 소속된 확률밀도가 서로 겹쳐진 형태의 이질군집 지역에 더 많은 코드벡터를 할당하기 위해서 본 논문에서 새로 제안이 제안되는 이질성 거리척도를 사용 하였다. 제안된 알고리즘을 한국어 고립 숫자단어의 인식문제에 적용한 결과, 기존 K-means 알고리즘이나 CNN보다 각각 14.63%, 9,39%의 오인식률의 감소를 얻을 수 있었다.

• 로봇학회논문지
• /
• 제12권4호
• /
• pp.432-440
• /
• 2017

In this paper, we propose a jellyfish distribution recognition and monitoring system using a UAV (unmanned aerial vehicle). The UAV was designed to satisfy the requirements for flight in ocean environment. The target jellyfish, Aurelia aurita, is recognized through convolutional neural network and its distribution is calculated. The modified deep neural network architecture has been developed to have reliable recognition accuracy and fast operation speed. Recognition speed is about 400 times faster than GoogLeNet by using a lightweight network architecture. We also introduce the method for selecting candidates to be used as inputs to the proposed network. The recognition accuracy of the jellyfish is improved by removing the probability value of the meaningless class among the probability vectors of the evaluated input image and re-evaluating it by normalization. The jellyfish distribution is calculated based on the unit jellyfish image recognized. The distribution level is defined by using the novelty concept of the distribution map buffer.

• Communications for Statistical Applications and Methods
• /
• 제7권2호
• /
• pp.605-616
• /
• 2000

Three primary interests frequently raised by mortgage companies are introduced and the corresponding statistical approaches for the default probability in mortgage companies are examined. Statistical models considered in this paper are time series, logistic regression, decision tree, neural network, and discrete time models. Usage of the models is illustrated using an artificially modified data set and the corresponding models are evaluated in appropriate manners.