• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.1
• /
• pp.672-679
• /
• 2018

This paper describes a methodology for shape design using an optimal design system, whereas generally a three dimensional analysis is required for such designs. An automatic finite element mesh generation technique, which is based on fuzzy knowledge processing and computational geometry techniques, is incorporated into the system, together with a commercial FE analysis code and a commercial solid modeler. Also, with the aid of multilayer neural networks, the present system allows us to automatically obtain a design window, in which a number of satisfactory design solutions exist in a multi-dimensional design parameter space. The developed optimal design system is successfully applied to evaluate the structures that are used. This study used a stress gauge to measure the maximum stress affecting the parts of the side housing bracket which are most vulnerable to cracking. Thereafter, we used a tool to interpret the maximum stress value, while maintaining the same stress as that exerted on the spot. Furthermore, a stress analysis was performed with the typical shape maintained intact, SM490 used for the material and the minimizing weight safety coefficient set to 3, while keeping the maximum stress the same as or smaller than the allowable stress. In this paper, a side housing bracket with a comparably simple structure for 36 tons was optimized, however if the method developed in this study were applied to side housing brackets of different classes (tons), their quality would be greatly improved.

• Journal of the Korea Convergence Society
• /
• v.12 no.10
• /
• pp.63-70
• /
• 2021

Generating a super-resolution meteological data by using a high-resolution deep neural network can provide precise research and useful real-life services. We propose a new technique of generating improved training data for super-resolution deep neural networks. To generate high-resolution meteorological data with domain specific knowledge, Lambert conformal conic projection and objective analysis were applied based on observation data and ERA5 reanalysis field data of specialized institutions. As a result, temperature and humidity analysis data based on domain specific knowledge showed improved RMSE by up to 42% and 46%, respectively. Next, a super-resolution generative adversarial network (SRGAN) which is one of the aritifial intelligence techniques was used to automate the manual data generation technique using damain specific techniques as described above. Experiments were conducted to generate high-resolution data with 1 km resolution from global model data with 10 km resolution. Finally, the results generated with SRGAN have a higher resoltuion than the global model input data, and showed a similar analysis pattern to the manually generated high-resolution analysis data, but also showed a smooth boundary.

• The Journal of Society for e-Business Studies
• /
• v.25 no.1
• /
• pp.13-33
• /
• 2020

This study proposes a Korean sentimental analysis algorithm that utilizes a letter-unit embedding and convolutional neural networks. Sentimental analysis is a natural language processing technique for subjective data analysis, such as a person's attitude, opinion, and propensity, as shown in the text. Recently, Korean sentimental analysis research has been steadily increased. However, it has failed to use a general-purpose sentimental dictionary and has built-up and used its own sentimental dictionary in each field. The problem with this phenomenon is that it does not conform to the characteristics of Korean. In this study, we have developed a model for analyzing emotions by producing syllable vectors based on the onset, peak, and coda, excluding morphology analysis during the emotional analysis procedure. As a result, we were able to minimize the problem of word learning and the problem of unregistered words, and the accuracy of the model was 88%. The model is less influenced by the unstructured nature of the input data and allows for polarized classification according to the context of the text. We hope that through this developed model will be easier for non-experts who wish to perform Korean sentimental analysis.

• Science of Emotion and Sensibility
• /
• v.12 no.1
• /
• pp.77-86
• /
• 2009

A comfort evaluation of car seat clothing has been proposed for high comforts interior seat clothing. Car seat covers have received wide spread attention due to their man-machine interface working. And then, it will be necessary for measurements on delicate basic mechanical-properties, which closely relate with human touch feeling of its materials. In this research, we have utilized $KES-FB^{(R)}$(Kawabata Evaluation System) series, $^ST300{(R)}$ analogue softness tester and friction tester for measurement a physical properties. In order to consider both kansei and physical properties on interior seat covers, we firstly have established subjective words of judgement for the seat covers. Secondly, related them to the objective measurement of physical properties. Each kansei-language has clearly defined as 'Softness', 'Elasticity', 'Volume' and 'Stickiness' for the adjectives of leather car seat covers. These technical terms have correlated to physical properties in other words, h (mm), bending moment ($gf^*$cm/cm), To-Tm (mm) and ${\mu}$. At this time, fuzzy logic has utilized to predict the value of kansei language through physical values. On the basis of this result, finally it is possible to predict quality index of car seat covers using neural networks technique. In short, we develop a quality evaluation system of car seat clothing combining four physical quantities with kansei engineering.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.34 no.6
• /
• pp.315-324
• /
• 2022

The experimental data obtained in a wave flume were analyzed using machine learning techniques to establish a model that predicts the input wave height of the wavemaker based on the waves that have experienced wave shoaling and to verify the performance of the established model. For this purpose, artificial neural network (NN), the most representative machine learning technique, and Gaussian process regression (GPR), one of the non-parametric regression analysis methods, were applied respectively. Then, the predictive performance of the two models was compared. The analysis was performed independently for the case of using all the data at once and for the case by classifying the data with a criterion related to the occurrence of wave breaking. When the data were not classified, the error between the input wave height at the wavemaker and the measured value was relatively large for both the NN and GPR models. On the other hand, if the data were divided into non-breaking and breaking conditions, the accuracy of predicting the input wave height was greatly improved. Among the two models, the overall performance of the GPR model was better than that of the NN model.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.49 no.2
• /
• pp.7-12
• /
• 2012

Recently the medical field to efficiently process the vast amounts of information to decision trees, neural networks, Bayesian Networks, including the application method of various data mining techniques are investigated. In addition, the basic personal information or patient history, family history, in addition to information such as MRI, HRCT images and additional information to collect and leverage in the diagnosis of disease, improved diagnostic accuracy is to promote a common status. But in real world situations that affect the results much because of the variable exists for a particular data mining techniques to obtain information through the enemy can be seen fairly limited. Medical images were taken as well as a minor can not give a positive impact on the diagnosis, but the proportion increased subjective judgments by the automated system is to deal with difficult issues. As a result of a complex reality, the situation is more advantageous to deal with the relative probability of the multivariate model based on Bayesian network, or TAN in the K2 search algorithm improves due to expansion model has been proposed. At this point, depending on the type of search algorithm applied significantly influenced the performance characteristics of the extended Bayesian network, the performance and suitability of each technique for evaluation of the facts is required. In this paper, we extend the Bayesian network for diagnosis of diseases using the same data were carried out, K2, TAN and changes in search algorithms such as classification accuracy was measured. In the 10-fold cross-validation experiment was performed to compare the performance evaluation based on the analysis and the onset of high-risk classification for patients with HRCT images could be possible to identify high-risk data.

• Korean Journal of Remote Sensing
• /
• v.33 no.4
• /
• pp.423-436
• /
• 2017

This paper presents a deep learning-based road segmentation framework from very high-resolution orthophotos. The proposed method uses Deep Convolutional Autoencoders for end-to-end mapping of orthophotos to road segmentations. In addition, a set of post-processing steps were applied to make the model outputs GIS-ready data that could be useful for various applications. The optimization of the model's parameters is explained which was conducted via grid search method. The model was trained and implemented in Keras, a high-level deep learning framework run on top of Tensorflow. The results show that the proposed model with the best-obtained hyperparameters could segment road objects from orthophotos at an average accuracy of 88.5%. The results of optimization revealed that the best optimization algorithm and activation function for the studied task are Stochastic Gradient Descent (SGD) and Exponential Linear Unit (ELU), respectively. In addition, the best numbers of convolutional filters were found to be 8 for the first and second layers and 128 for the third and fourth layers of the proposed network architecture. Moreover, the analysis on the time complexity of the model showed that the model could be trained in 4 hours and 50 minutes on 1024 high-resolution images of size $106{\times}106pixels$, and segment road objects from similar size and resolution images in around 14 minutes. The results show that the deep learning models such as Convolutional Autoencoders could be a best alternative to traditional machine learning models for road segmentation from aerial photographs.

• The Korean Journal of Physiology and Pharmacology
• /
• v.22 no.1
• /
• pp.71-80
• /
• 2018

In patients with epilepsy, depression is a common comorbidity but difficult to be treated because many antidepressants cause pro-convulsive effects. Thus, it is important to identify the risk of seizures associated with antidepressants. To determine whether paroxetine, a very potent selective serotonin reuptake inhibitor (SSRI), interacts with ion channels that modulate neuronal excitability, we examined the effects of paroxetine on Kv3.1 potassium channels, which contribute to high-frequency firing of interneurons, using the whole-cell patch-clamp technique. Kv3.1 channels were cloned from rat neurons and expressed in Chinese hamster ovary cells. Paroxetine reversibly reduced the amplitude of Kv3.1 current, with an $IC_{50}$ value of $9.43{\mu}M$ and a Hill coefficient of 1.43, and also accelerated the decay of Kv3.1 current. The paroxetine-induced inhibition of Kv3.1 channels was voltage-dependent even when the channels were fully open. The binding ($k_{+1}$) and unbinding ($k_{-1}$) rate constants for the paroxetine effect were $4.5{\mu}M^{-1}s^{-1}$ and $35.8s^{-1}$, respectively, yielding a calculated $K_D$ value of $7.9{\mu}M$. The analyses of Kv3.1 tail current indicated that paroxetine did not affect ion selectivity and slowed its deactivation time course, resulting in a tail crossover phenomenon. Paroxetine inhibited Kv3.1 channels in a use-dependent manner. Taken together, these results suggest that paroxetine blocks the open state of Kv3.1 channels. Given the role of Kv3.1 in fast spiking of interneurons, our data imply that the blockade of Kv3.1 by paroxetine might elevate epileptic activity of neural networks by interfering with repetitive firing of inhibitory neurons.

• Proceedings of the Korea Society of Environmental Toocicology Conference
• /
• 2002.10a
• /
• pp.156-156
• /
• 2002

The goal of this study is to develop a biomarker used in monitoring abnormal behaviors of Japanese medaka (Oryzias latipes) as a model organism caused by hazardous chemicals that are toxic and persistent in the ecosystem. A widely used insecticide, diazinon (O, O-diethyl O- (2-isopropyl-4-methyl-6-pyrimidinyl) phosphorothioate), is highly neurotoxic to fish, and it is also well known that it causes vertebral malformation and behavioral changes of fish at relatively low concentrations. The fish behaviors were observed on a real time basis using an image processing and automatic data acquisition system. The genes potentially involved in the abnormal behaviors were cloned using suppression subtractive hybridization (SSH) technique. The untreated individuals showed common behavioral characteristics. When the test fish was affected by diazinon at a concentration of 0.1 and 1 ppm, some specific patterns were observed in its behavioral activity and locomotive tracks. The typical patterns were enhanced surfacing activity, opercular movement, erratic movement, tremors and convulsions as reported previously. The number of genes up-regulated tty diazinon treatment were 97 which includes 27 of unknown genes. The number of down-regulated genes were 99 including 60 of unknown genes. These gene expression patterns will be analyzed by the artificial neural networks such as self organization map (SOM) and multilayer perceptron (MLP), revealing the role of genes responsible for the behaviors. These results may provide molecular biological and neurobehavioral bases of a biomonitoring system for diazinon using a model organism such as fish.

• Journal of Digital Contents Society
• /
• v.12 no.1
• /
• pp.1-9
• /
• 2011

JPEG2000 offers the technique which compresses the interested regions with higher quality than the background. It is called by an ROI(Region-of-Interest) coding method. In this paper, we use images including the human faces, which are processed uppermost and compressed with high quality. The proposed method consists of 2 steps. The first step extracts some faces and the second one is ROI coding. To extract the faces, the method cuts or scale-downs some regions with $20{\times}20$ window pixels for all the pixels of the image, and after preprocessing, recognizes the faces using neural networks. Each extracted region is identified by ROI mask and then ROI-coded using Maxshift method. After then, the image is compressed and saved using EBCOT. The existing methods searched the ROI by edge distributions. On the contrary, the proposed method uses human intellect. And the experiment shows that the method is sufficiently useful with images having several human faces.