• Geophysics and Geophysical Exploration
• /
• v.14 no.3
• /
• pp.220-226
• /
• 2011

In the elastic wave equations, both horizontal and vertical displacements are defined. Since we can measure both the horizontal and vertical displacements in field acquisition, these displacements compose a displacement vector. In this study, we propose a frequency-domain elastic waveform inversion technique taking advantage of the magnitudes of displacement vectors to define objective function. When we apply this displacement-vector objective function to the frequency-domain waveform inversion, the inversion process naturally incorporates the back-propagation algorithm. Through the inversion examples with the Marmousi model and the SEG/EAGE salt model, we could note that the RMS error of the solution obtained by our algorithm decreased more stably than that of the conventional method. Particularly, the density of the Marmousi model and the low-velocity sub-salt zone of the SEG/EAGE salt model were successfully recovered. Since the gradient direction obtained from the proposed objective function is numerically unstable, we need additional study to stabilize the gradient direction. In order to perform the waveform inversion using the displacementvector objective function, it is necessary to acquire multi-component data. Hence, more rigorous study should be continued for the multi-component land acquisition or OBC (Ocean Bottom Cable) multi-component survey.

• Journal of Intelligence and Information Systems
• /
• v.22 no.2
• /
• pp.127-142
• /
• 2016

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for "backward propagation of errors" and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer's) neurons. Shared weights mean that we're going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren't just propagated backward through layers, they're propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when training RNNs, and many recent papers make use of LSTMs or related ideas.

• Journal of radiological science and technology
• /
• v.26 no.1
• /
• pp.71-82
• /
• 2003

In this paper, I proposed a classifier of liver cirrhotic step using T1-weighted MRI(magnetic resonance imaging) and hierarchical neural network. The data sets for classification of each stage, which were normal, 1type, 2type and 3type, were obtained in Pusan National University Hospital from June 2001 to december 2001. And the number of data was 46. We extracted liver region and nodule region from T1-weighted MR liver image. Then objective interpretation classifier of liver cirrhotic steps in T1-weighted MR liver images. Liver cirrhosis classifier implemented using hierarchical neural network which gray-level analysis and texture feature descriptors to distinguish normal liver and 3 types of liver cirrhosis. Then proposed Neural network classifier teamed through error back-propagation algorithm. A classifying result shows that recognition rate of normal is 100%, 1type is 82.3%, 2type is 86.7%, 3type is 83.7%. The recognition ratio very high, when compared between the result of obtained quantified data to that of doctors decision data and neural network classifier value. If enough data is offered and other parameter is considered, this paper according to we expected that neural network as well as human experts and could be useful as clinical decision support tool for liver cirrhosis patients.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.42 no.2 s.332
• /
• pp.61-74
• /
• 2005

Digital Multimedia Broadcasting(DMB) is a reliable multi-service system for reception by mobile and portable receivers. DMB system allows interference-free reception under the conditions of multipath propagation and transmission errors using COFDM modulation scheme, simultaneously, needs powerful channel error's correction ability. Viterbi Decoder for DMB receiver uses punctured convolutional code and needs lots of computations for real-time operation. So, it is desired to design a high speed and low-power hardware scheme for Viterbi decoder. This paper proposes a combined add-compare-select(ACS) and path metric normalization(PMN) unit for computation power. The proposed PMN architecture reduces the problem of the critical path by applying fixed value for selection algorithm due to the comparison tree which has a weak point from structure with the high-speed operation. The proposed ACS uses the decomposition and the pre-computation technique for reducing the complicated degree of the adder, the comparator and multiplexer. According to a simulation result, reduction of area $3.78\%$, power consumption $12.22\%$, maximum gate delay $23.80\%$ occurred from punctured viterbi decoder for DMB system.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.10 no.5
• /
• pp.487-496
• /
• 2000

In this paper, the Multi-FNN(Fuzzy-Neural Networks) model is identified and optimized using HCM(Hard C-Means) clustering method and genetic algorithms. The proposed Multi-FNN is based on Yamakawa's FNN and uses simplified inference as fuzzy inference method and error back propagation algorithm as learning rules. We use a HCM clustering and Genetic Algorithms(GAs) to identify both the structure and the parameters of a Multi-FNN model. Here, HCM clustering method, which is carried out for the process data preprocessing of system modeling, is utilized to determine the structure of Multi-FNN according to the divisions of input-output space using I/O process data. Also, the parameters of Multi-FNN model such as apexes of membership function, learning rates and momentum coefficients are adjusted using genetic algorithms. A aggregate performance index with a weighting factor is used to achieve a sound balance between approximation and generalization abilities of the model. The aggregate performance index stands for an aggregate objective function with a weighting factor to consider a mutual balance and dependency between approximation and predictive abilities. According to the selection and adjustment of a weighting factor of this aggregate abjective function which depends on the number of data and a certain degree of nonlinearity, we show that it is available and effective to design an optimal Multi-FNN model. To evaluate the performance of the proposed model, we use the time series data for gas furnace and the numerical data of nonlinear function.

• Journal of Biomedical Engineering Research
• /
• v.23 no.3
• /
• pp.197-205
• /
• 2002

The objectives of this Paper is to implement a diagnostic classifier of differential laryngeal diseases from acoustic signals acquired in a noisy room. For this Purpose, the voice signals of the vowel /a/ were collected from Patients in a soundproof chamber and got mixed with noise. Then, the acoustic Parameters were analyzed, and hierarchical neural networks were applied to the data classification. The classifier had a structure of five-step hierarchical neural networks. The first neural network classified the group into normal and benign or malign laryngeal disease cases. The second network classified the group into normal or benign laryngeal disease cases The following network distinguished polyp. nodule. Palsy from the benign laryngeal cases. Glottic cancer cases were discriminated into T1, T2. T3, T4 by the fourth and fifth networks All the neural networks were based on multilayer perceptron model which classified non-linear Patterns effectively and learned by an error back-propagation algorithm. We chose some acoustic Parameters for classification by investigating the distribution of laryngeal diseases and Pilot classification results of those Parameters derived from MDVP. The classifier was tested by using the chosen parameters to find the optimum ones. Then the networks were improved by including such Pre-Processing steps as linear and z-score transformation. Results showed that 90% of T1, 100% of T2-4 were correctly distinguished. On the other hand. 88.23% of vocal Polyps, 100% of normal cases. vocal nodules. and vocal cord Paralysis were classified from the data collected in a noisy room.

• Journal of Korea Water Resources Association
• /
• v.31 no.1
• /
• pp.13-25
• /
• 1998

It is necessary to develop methodologies for the application of artificial neural network into hydrologic rainfall-runoff process, although there is so much applicability by using the functions of associative memory based on recognition for the relationships between causes and effects and the excellent fitting capacity for the nonlinear phenomenon. In this study, some problems are presented in the application procedures of artificial neural networks and the simulation of runoff hydrograph experiences are reviewed with nonlinear functional approximator by artificial neural network for rainfall-runoff relationships in a watershed. which is regarded as hydrdologic black box model. The neural network models are constructed by organizing input and output patterns with the deserved rainfall and runoff data in Pyoungchang river basin under the assumption that the rainfall data is the input pattern and runoff hydrograph is the output patterns. Analyzed with the results. it is possible to simulate the runoff hydrograph with processing element of artificial neural network with any hydrologic concepts and the weight among processing elements are well-adapted as model parameters with the assumed model structure during learning process. Based upon these results. it is expected that neural network theory can be utilized as an efficient approach to simulate runoff hydrograph and identify the relationship between rainfall and runoff as hydrosystems which is necessary to develop and manage water resources.

• Journal of Intelligence and Information Systems
• /
• v.19 no.2
• /
• pp.73-85
• /
• 2013

Nowadays, in today's information society, the importance of the knowledge service using the information to creative value is getting higher day by day. In addition, depending on the development of IT technology, it is ease to collect and use information. Also, many companies actively use customer information to marketing in a variety of industries. Into the 21st century, companies have been actively using the culture arts to manage corporate image and marketing closely linked to their commercial interests. But, it is difficult that companies attract or maintain consumer's interest through their technology. For that reason, it is trend to perform cultural activities for tool of differentiation over many firms. Many firms used the customer's experience to new marketing strategy in order to effectively respond to competitive market. Accordingly, it is emerging rapidly that the necessity of personalized service to provide a new experience for people based on the personal profile information that contains the characteristics of the individual. Like this, personalized service using customer's individual profile information such as language, symbols, behavior, and emotions is very important today. Through this, we will be able to judge interaction between people and content and to maximize customer's experience and satisfaction. There are various relative works provide customer-centered service. Specially, emotion recognition research is emerging recently. Existing researches experienced emotion recognition using mostly bio-signal. Most of researches are voice and face studies that have great emotional changes. However, there are several difficulties to predict people's emotion caused by limitation of equipment and service environments. So, in this paper, we develop emotion prediction model based on vision-based interface to overcome existing limitations. Emotion recognition research based on people's gesture and posture has been processed by several researchers. This paper developed a model that recognizes people's emotional states through body gesture and posture using difference image method. And we found optimization validation model for four kinds of emotions' prediction. A proposed model purposed to automatically determine and predict 4 human emotions (Sadness, Surprise, Joy, and Disgust). To build up the model, event booth was installed in the KOCCA's lobby and we provided some proper stimulative movie to collect their body gesture and posture as the change of emotions. And then, we extracted body movements using difference image method. And we revised people data to build proposed model through neural network. The proposed model for emotion prediction used 3 type time-frame sets (20 frames, 30 frames, and 40 frames). And then, we adopted the model which has best performance compared with other models.' Before build three kinds of models, the entire 97 data set were divided into three data sets of learning, test, and validation set. The proposed model for emotion prediction was constructed using artificial neural network. In this paper, we used the back-propagation algorithm as a learning method, and set learning rate to 10%, momentum rate to 10%. The sigmoid function was used as the transform function. And we designed a three-layer perceptron neural network with one hidden layer and four output nodes. Based on the test data set, the learning for this research model was stopped when it reaches 50000 after reaching the minimum error in order to explore the point of learning. We finally processed each model's accuracy and found best model to predict each emotions. The result showed prediction accuracy 100% from sadness, and 96% from joy prediction in 20 frames set model. And 88% from surprise, and 98% from disgust in 30 frames set model. The findings of our research are expected to be useful to provide effective algorithm for personalized service in various industries such as advertisement, exhibition, performance, etc.