• IE interfaces
• /
• v.10 no.2
• /
• pp.51-56
• /
• 1997

This paper concerns with the problem of mixed model assembly sequencing using neural net. In recent years, because of two characteristics of it, massive parallelism and learning capability, neural nets have emerged to solve the problems for which more conventional computational approaches have proven ineffective. This paper proposes a method using neural net that can consider line balancing and grouping problems simultaneously. In order to solve the mixed model assembly sequencing of the motor industry, this paper uses the modified ART1 algorithm.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1993.06a
• /
• pp.1326-1329
• /
• 1993

This work presents a net-based structure to model approximate reasoning using fuzzy production rules, the Fuzzy Petri Net model. The Fuzzy Petri Net model is formally defined as a n-uple of elements. It allows for the representation of simple and complex forms of rules such as rules with conjunction in the antecedent and qualified rules. Parallel rules and conflicting rules can be modeled as well. We also developed an analysis method based on state equations and two fuzzy reasoning algorithms. Finally, the proposed method is applied to an example.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2004.05a
• /
• pp.243-246
• /
• 2004

본 논문은 영상인식 방법으로 개선된 Active Model을 이용한 방법을 제안한다. 제안된 방법은 모든 격자 블록 영역이 동일한 구조를 가지며, 기존의 Active net에서 문제가 되었던 목표물을 탐지하는 능력이 개선되었다. 실험 결과로서 제안된 방법이 수직, 수평 방향에서 목표물 포착에 효과적임을 보여주었으며, 실제 도로 영상에 적용한 결과 제안한 방법의 효율성을 입증할 수 있었다.

• The Journal of Industrial Distribution & Business
• /
• v.12 no.2
• /
• pp.41-52
• /
• 2021

Purpose: This study aims to provide implications for the government's housing supply policy by analyzing the factors that determine the type of real estate holding and household debt. This study started from the awareness that the determinants of household debt differ depending on the type of real estate holding. Research design, data and methodology: Real estate ownership type was classified and analyzed into 4 models: model 1 (1 household 1 house and self-resident), model 2 (1 household multiple real estate ownership and self-resident), model 3 (1 household 1 house and rent residence), model 4 (1 household holds a large number of real estate and rent residence). The analysis method used multiple regression analysis. The dependent variable was household total debt. As independent variables, household debt, annual gross household income, financial assets, real estate net assets, annual repayment, demographic & residential characteristics were used. Results: 1) Model 4 has the highest household debt and the highest gross income, Model 2 has the most real estate mortgage loans and real estate net asset, and Model 1 has the highest real estate mortgage payments. 2) The positive factor of common household debt determinants is real estate net assets, and the negative factor is financial assets. 3) It was the net assets of real estate that acted as a positive factor in common for the four models. In other words, the more financial assets, the less household debt. It was analyzed that the more net assets of real estate, the more household debt. The annual repayment of financial liabilities had no influence on household debt, while the annual repayment of loan liabilities and household debt had a positive relationship. Conclusions: 1) It is necessary to introduce benefits and systems that can increase the proportion of household financial asset. Specific alternatives include tax benefits and reduced fees for financial asset investment. 2) In the case where a homeless person prepares one house for one household, it is necessary to prepare various support measures according to the income level. The specific alternative is to give additional points for pre-sale or apply an interest rate cut incentive for mortgage loans.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.25 no.2
• /
• pp.26-38
• /
• 2021

Image coloration refers to adding plausible colors to a grayscale image or video. Image coloration has been used in many modern fields, including restoring old photographs, as well as reducing the time spent painting cartoons. In this paper, a method is proposed for colorizing grayscale images using a convolutional neural network. We propose an encoder-decoder model, adapting FusionNet to our purpose. A proper loss function is defined instead of the MSE loss function to suit the purpose of coloring. The proposed model was verified using the ImageNet dataset. We quantitatively compared several colorization models with ours, using the peak signal-to-noise ratio (PSNR) metric. In addition, to qualitatively evaluate the results, our model was applied to images in the test dataset and compared to images applied to various other models. Finally, we applied our model to a selection of old black and white photographs.

• Ocean and Polar Research
• /
• v.37 no.2
• /
• pp.161-177
• /
• 2015

This study analyzes the stock assessment of yellow croaker caught mainly by the Korean stow net and gill net fisheries focusing on single species and multiple fisheries. This study standardizes fishing efforts for the two fisheries using the general linear model and uses a surplus production model based on the exponential growth model. The Clarke Yoshimoto Pooley model estimates a maximum sustainable yield(MSY), an allowable biological catch(ABC), fishing efforts for MSY($E_{MSY}$) and for ABC($E_{ABC}$). The bio-economic model is used to estimate the maximum economic yield(MEY) and fishing efforts for MEY($E_{MSY}$). Also, the study employs an economic analysis to estimate the economic interaction between stow net and gill net fisheries. The economic analysis shows the profit accruing to the two fisheries from estimated ABC. Finally, the study compares TACs based on single species and single fishery to TAC based on single species and multiple fisheries. The study proposes that the TAC assessment is necessary for single species and multiple fisheries in order to preserve resources.

• International Journal of Computer Science & Network Security
• /
• v.22 no.7
• /
• pp.301-307
• /
• 2022

Energy consumption has grown alongside dramatic population increases. Statistics show that buildings in particular utilize a significant amount of energy, worldwide. Because of this, building energy prediction is crucial to best optimize utilities' energy plans and also create a predictive model for consumers. To improve energy prediction performance, this paper proposes a ResNet-LSTM model that combines residual networks (ResNets) and long short-term memory (LSTM) for energy consumption prediction. ResNets are utilized to extract complex and rich features, while LSTM has the ability to learn temporal correlation; the dense layer is used as a regression to forecast energy consumption. To make our model more robust, we employed Huber loss during the optimization process. Huber loss obtains high efficiency by handling minor errors quadratically. It also takes the absolute error for large errors to increase robustness. This makes our model less sensitive to outlier data. Our proposed system was trained on historical data to forecast energy consumption for different time series. To evaluate our proposed model, we compared our model's performance with several popular machine learning and deep learning methods such as linear regression, neural networks, decision tree, and convolutional neural networks, etc. The results show that our proposed model predicted energy consumption most accurately.

• International Journal of Computer Science & Network Security
• /
• v.23 no.5
• /
• pp.73-88
• /
• 2023

Brain tumors can also be an abnormal collection or accumulation of cells in the brain that can be life-threatening due to their ability to invade and metastasize to nearby tissues. Accurate diagnosis is critical to the success of treatment planning, and resonant imaging is the primary diagnostic imaging method used to diagnose brain tumors and their extent. Deep learning methods for computer vision applications have shown significant improvements in recent years, primarily due to the undeniable fact that there is a large amount of data on the market to teach models. Therefore, improvements within the model architecture perform better approximations in the monitored configuration. Tumor classification using these deep learning techniques has made great strides by providing reliable, annotated open data sets. Reduce computational effort and learn specific spatial and temporal relationships. This white paper describes transfer models such as the MobileNet model, VGG19 model, InceptionResNetV2 model, Inception model, and DenseNet201 model. The model uses three different optimizers, Adam, SGD, and RMSprop. Finally, the pre-trained MobileNet with RMSprop optimizer is the best model in this paper, with 0.995 accuracies, 0.99 sensitivity, and 1.00 specificity, while at the same time having the lowest computational cost.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.43 no.3
• /
• pp.192-205
• /
• 2007

The objective of this study was to investigate the optimal shapes and arrangements of sinkers attached to net cages to prevent their deformation in a current. A series of model experiments were conducted in a circulating water channel, using 5 different types of sinker(high-weighted ball, low-weighted ball, columntype, egg-shaped and iron bar-framed) and 2 types of square net cage constructed from both Nylon Raschel netting and Nylon knotted netting, on a 1/20th scale. The deflection of the model nets against the flow was smallest with the iron bar-framed weight compared to the other four types of sinker. It was expected that the optimal shapes of sinkers would be either the ball or egg-shape; however, iron bar-framed weight actually had larger drag forces. The dispersed deployment of sinkers on the bottom frames of model net cages performed better with relatively slow flows, while the concentrated deployment at 4 corners functioned better with relatively fast flows, in preventing the nets from becoming severely deformed. The deformation of the net cages was larger for the Nylon knotted netting than the Nylon Raschel netting. With respect to flow resistance, the Nylon Raschel netting, rather than the Nylon knotted netting, was more suitable for construction of net cages.

• Smart Media Journal
• /
• v.11 no.11
• /
• pp.92-98
• /
• 2022

Identifying defects in textiles is a key procedure for quality control. This study attempted to create a model that detects defects by analyzing the images of the fabrics. The models used in the study were deep learning-based VGGNet and ResNet, and the defect detection performance of the two models was compared and evaluated. The accuracy of the VGGNet and the ResNet model was 0.859 and 0.893, respectively, which showed the higher accuracy of the ResNet. In addition, the region of attention of the model was derived by using the Grad-CAM algorithm, an eXplainable Artificial Intelligence (XAI) technique, to find out the location of the region that the deep learning model recognized as a defect in the fabric image. As a result, it was confirmed that the region recognized by the deep learning model as a defect in the fabric was actually defective even with the naked eyes. The results of this study are expected to reduce the time and cost incurred in the fabric production process by utilizing deep learning-based artificial intelligence in the defect detection of the textile industry.