• Journal of Navigation and Port Research
• /
• v.43 no.1
• /
• pp.72-78
• /
• 2019

The second-hand ship market provides immediate access to the freight market for shipping investors. When introducing second-hand vessels, the precise estimate of the price is crucial to the decision-making process because it directly affects the burden of capital cost to investors in the future. Previous studies on the second-hand market have mainly focused on the market efficiency. The number of papers on the estimation of second-hand vessel values is very limited. This study proposes an artificial neural network model that has not been attempted in previous studies. Six factors, freight, new-building price, orderbook, scrap price, age and vessel size, that affect the second-hand ship price were identified through literature review. The employed data is 366 real trading records of Panamax second-hand vessels reported to Clarkson between January 2016 and December 2018. Statistical filtering was carried out through correlation analysis and stepwise regression analysis, and three parameters, which are freight, age and size, were selected. Ten-fold cross validation was used to estimate the hyper-parameters of the artificial neural network model. The result of this study confirmed that the performance of the artificial neural network model is better than that of simple stepwise regression analysis. The application of the statistical verification process and artificial neural network model differentiates this paper from others. In addition, it is expected that a scientific model that satisfies both statistical rationality and accuracy of the results will make a contribution to real-life practices.

• Journal of Broadcast Engineering
• /
• v.21 no.1
• /
• pp.36-42
• /
• 2016

Recently, many 3D contents have been produced through the multiview camera system. In this system, since a difference of the viewpoint between color and depth cameras is inevitable, the camera parameter plays the important role to adjust the viewpoint as a preprocessing step. The conventional camera calibration method is inconvenient to users since we need to choose pattern features manually after capturing a planar chessboard with various poses. Therefore, we propose a semi-auto camera calibration method using a circular sampling and an homography estimation. Firstly, The proposed method extracts the candidates of the pattern features from the images by FAST corner detector. Next, we reduce the amount of the candidates by the circular sampling and obtain the complete point cloud by the homography estimation. Lastly, we compute the accurate position having the sub-pixel accuracy of the pattern features by the approximation of the hyper parabola surface. We investigated which factor affects the result of the pattern feature detection at each step. Compared to the conventional method, we found the proposed method released the inconvenience of the manual operation but maintained the accuracy of the camera parameters.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.30 no.4
• /
• pp.617-629
• /
• 2020

In order to overcome the limitations of the rule-based intrusion detection system due to changes in Internet computing environments, the emergence of new services, and creativity of attackers, network anomaly detection (NAD) using machine learning and deep learning technologies has received much attention. Most of these existing machine learning and deep learning technologies for NAD use supervised learning methods to learn a set of training data set labeled 'normal' and 'attack'. This paper presents the feasibility of the unsupervised learning AutoEncoder(AE) to NAD from data sets collecting of secured network traffic without labeled responses. To verify the performance of the proposed AE mode, we present the experimental results in terms of accuracy, precision, recall, f1-score, and ROC AUC value on the NSL-KDD training and test data sets. In particular, we model a reference AE through the deep analysis of diverse AEs varying hyper-parameters such as the number of layers as well as considering the regularization and denoising effects. The reference model shows the f1-scores 90.4% and 89% of binary classification on the KDDTest+ and KDDTest-21 test data sets based on the threshold of the 82-th percentile of the AE reconstruction error of the training data set.

• Clinical and Experimental Reproductive Medicine
• /
• v.16 no.2
• /
• pp.205-210
• /
• 1989

To investigate the effects of ovarian cysts on the controlled ovarian hyper-stimulation cycles, 16 patients with 16 paired cycles for IVF-ET were analyzed. These patients had taken both type of cycles, i.e., with cyst(cyst group) and without cyst(control group). Mean diameter of ovarian cysts in cyst group was 18.2mm. There were no significant differences in hormone levels in early follicular phase between two groups. No significant differences were found in total dosage of hMG(IU) administered during the ovarian stimulation $843.8{\pm}123.0$ vs $891.0{\pm}129.8$, serum estradiol level (pg/ml) on the day of hCG administration($1542.8{\pm}1100.6$ vs $1567.5{\pm}1193.0$), the number of aspirated follicles $10.0{\pm}3.4$ vs $11.2{\pm}4.3$ and oocytes $5.3{\pm}3.3$ vs $6.2{\pm}3.1$, the fertilization rate(51.2 % vs 57.2 %) and the cleavage rate(40.5 % vs 52.0 %). Serum estradiol terminal patterns during COH in one group tended to be repeated in the other group. In conclusion, this study suggests that small ovarian cysts do not adversely impact on the controlled ovarian hyperstimulation parameters in IVF - ET program and the presence of small ovarian cyst without concomitant high basal serum estradiol level is not an indication of the cancellation of the controlled ovarian hyperstimulation for IVF-ET.

• Journal of Digital Convergence
• /
• v.18 no.11
• /
• pp.259-266
• /
• 2020

Many developments have been steadily carried out by researchers with applying knowledge-based expert system or machine learning algorithms to the financial field. In particular, it is now common to perform knowledge based system trading in using stock prices. Recently, deep learning technologies have been applied to real fields of stock trading marketplace as GPU performance and large scaled data have been supported enough. Especially, LSTM has been tried to apply to stock price prediction because of its compatibility for time series data. In this paper, we implement stock price prediction using LSTM. In modeling of LSTM, we propose a fitness combination of model parameters and activation functions for best performance. Specifically, we propose suitable selection methods of initializers of weights and bias, regularizers to avoid over-fitting, activation functions and optimization methods. We also compare model performances according to the different selections of the above important modeling considering factors on the real-world stock price data of global major companies. Finally, our experimental work brings a fitness method of applying LSTM model to stock price prediction.

• Asian-Australasian Journal of Animal Sciences
• /
• v.21 no.6
• /
• pp.794-800
• /
• 2008

In the present study, two trials were conducted to evaluate the effects of hyper- and hypothyroid status on the redox balance of broiler chickens. In Trial 1, 3 groups of broiler chickens were randomly subjected to one of the three treatments: subcutaneous administration of triiodothyronine (T3, $150{\mu}g/kg$ BW), methimazole (MMI, 150 mg/kg BW), or saline. The blood, liver and heart were sampled at 3 h after injection. In Trial 2, three groups of 20 broiler chickens were randomly fed with one of the three diets: control, dietary supplementation of T3 (1.5 mg/kg diet) or MMI (1 g/kg diet) for 7 days. In trial 1, the plasma concentrations of T3 and T3 to thyronine ratio (T3/T4) were significantly increased by T3 injection. Plasma levels of thiobarbituric acid reacting substances (TBARS) tended to be increased (p = 0.067) by both T3 and MMI treatments while the ferric reduced/antioxidant capacity (FRAP) was increased only by MMI treatment. Acute T3 treatment had no significant effect on the activities of superoxide dismutase (SOD) and the concentrations of FRAP and TBARS in either liver or heart tissue. In contrast, the hepatic activities of SOD were decreased (p<0.05) while the cardiac levels of FRAP were significantly increased (p<0.0001) by MMI treatment. In chronic treatments, the rectal temperature of chickens was significantly decreased (p<0.05) by MMI treatment. The circulating T3 levels were significantly increased (p<0.05) by long-term T3 treatment, and showed a trend to decrease in MMI treatment. The plasma concentrations of TBARS were significantly (p<0.05) increased by MMI treatment. All the redox parameters measured in either liver or heart were not significantly altered by either long-term T3 or MMI treatment except that the hepatic SOD activities were significantly augmented by T3 treatment. The result showed that neither acute nor long-term elevation of circulating T3 levels induced lipid peroxidation in broiler chickens. The enhanced enzymatic antioxidant system (SOD in cardiac tissue) may be involved in the protection of the bird to increased oxidative challenge. The responses of redox balance to changed thyroid state seem to be tissue specific.

• Composites Research
• /
• v.33 no.3
• /
• pp.108-114
• /
• 2020

Piecewise Integrated Composite (PIC) beam is composed of different stacking against loading type depending upon location. The aim of current study is to assign robust stacking sequences against external loading to every corresponding part of the PIC beam based on the value of stress triaxiality at generated reference points using the k-NN (k-Nearest Neighbor) classification, which is one of representative machine learning techniques, in order to excellent superior bending characteristics. The stress triaxiality at reference points is obtained by three-point bending analysis of the Al beam with training data categorizing the type of external loading, i.e., tension, compression or shear. Loading types of each plane of the beam were classified by independent plane scheme as well as total beam scheme. Also, loading fidelities were calibrated for each case with the variation of hyper-parameters. Most effective stacking sequences were mapped into the PIC beam based on the k-NN classification model with the highest loading fidelity. FE analysis result shows the PIC beam has superior external loading resistance and energy absorption compared to conventional beam.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.34 no.5
• /
• pp.525-534
• /
• 2016

In South Korea, paddy rice has been consumed over the entire region and it is the main source of income for farmers, thus mathematical model for the estimation of rice yield is required for such decision-making processes in agriculture. The objectives of our study are to: (1) develop rice yield estimation model using Convolutional Neural Networks(CNN), (2) choose hyper-parameters for the model which show the best performance and (3) investigate whether CNN model can effectively predict the rice yield by the comparison with the model using Artificial Neural Networks(ANN). Weather and MODIS(The MOderate Resolution Imaging Spectroradiometer) products from April to September in year 2000~2013 were used for the rice yield estimation models and cross-validation was implemented for the accuracy assessment. The CNN and ANN models showed Root Mean Square Error(RMSE) of 36.10kg/10a, 48.61kg/10a based on rice points, respectively and 31.30kg/10a, 39.31kg/10a based on 'Si-Gun-Gu' districts, respectively. The CNN models outperformed ANN models and its possibility of application for the field of rice yield estimation in South Korea was proved.

• Korean Journal of Remote Sensing
• /
• v.35 no.5_1
• /
• pp.681-692
• /
• 2019

In this paper, a hybrid deep learning model, called 2D convolution with bidirectional long short-term memory (2DCBLSTM), is presented that can effectively combine both spatial and temporal features for crop classification. In the proposed model, 2D convolution operators are first applied to extract spatial features of crops and the extracted spatial features are then used as inputs for a bidirectional LSTM model that can effectively process temporal features. To evaluate the classification performance of the proposed model, a case study of crop classification was carried out using multi-temporal unmanned aerial vehicle images acquired in Anbandegi, Korea. For comparison purposes, we applied conventional deep learning models including two-dimensional convolutional neural network (CNN) using spatial features, LSTM using temporal features, and three-dimensional CNN using spatio-temporal features. Through the impact analysis of hyper-parameters on the classification performance, the use of both spatial and temporal features greatly reduced misclassification patterns of crops and the proposed hybrid model showed the best classification accuracy, compared to the conventional deep learning models that considered either spatial features or temporal features. Therefore, it is expected that the proposed model can be effectively applied to crop classification owing to its ability to consider spatio-temporal features of crops.

• Journal of Korea Water Resources Association
• /
• v.51 no.12
• /
• pp.1207-1216
• /
• 2018

Discharge or water level predictions at tidally affected river reaches are currently still a great challenge in hydrological practices. This research aims to predict water level of the tide dominated site, Jamsu bridge in the Han River downstream. Physics-based hydrodynamic approaches are sometimes not applicable for water level prediction in such a tidal river due to uncertainty sources like rainfall forecasting data. In this study, TensorFlow deep learning framework was used to build a deep neural network based LSTM model and its applications. The LSTM model was trained based on 3 data sets having 10-min temporal resolution: Paldang dam release, Jamsu bridge water level, predicted tidal level for 6 years (2011~2016) and then predict the water level time series given the six lead times: 1, 3, 6, 9, 12, 24 hours. The optimal hyper-parameters of LSTM model were set up as follows: 6 hidden layers number, 0.01 learning rate, 3000 iterations. In addition, we changed the key parameter of LSTM model, sequence length, ranging from 1 to 6 hours to test its affect to prediction results. The LSTM model with the 1 hr sequence length led to the best performing prediction results for the all cases. In particular, it resulted in very accurate prediction: RMSE (0.065 cm) and NSE (0.99) for the 1 hr lead time prediction case. However, as the lead time became longer, the RMSE increased from 0.08 m (1 hr lead time) to 0.28 m (24 hrs lead time) and the NSE decreased from 0.99 (1 hr lead time) to 0.74 (24 hrs lead time), respectively.