• Journal of Internet Computing and Services
• /
• v.22 no.1
• /
• pp.23-30
• /
• 2021

This paper proposes a noise-tolerant image classification system using multiple autoencoders. The development of deep learning technology has dramatically improved the performance of image classifiers. However, if the images are contaminated by noise, the performance degrades rapidly. Noise added to the image is inevitably generated in the process of obtaining and transmitting the image. Therefore, in order to use the classifier in a real environment, we have to deal with the noise. On the other hand, the autoencoder is an artificial neural network model that is trained to have similar input and output values. If the input data is similar to the training data, the error between the input data and output data of the autoencoder will be small. However, if the input data is not similar to the training data, the error will be large. The proposed system uses the relationship between the input data and the output data of the autoencoder, and it has two phases to classify the images. In the first phase, the classes with the highest likelihood of classification are selected and subject to the procedure again in the second phase. For the performance analysis of the proposed system, classification accuracy was tested on a Gaussian noise-contaminated MNIST dataset. As a result of the experiment, it was confirmed that the proposed system in the noisy environment has higher accuracy than the CNN-based classification technique.

• Health Policy and Management
• /
• v.30 no.4
• /
• pp.493-504
• /
• 2020

Background: There have been many studies on the associations between body mass index (BMI) and cognitive function. However, no study has ever compared the associations across the methods of categorizing BMI. In this study, we aimed to fill the gap in the previous studies and examine whether the obesity paradox is valid in the risk of cognitive function. Methods: Of the 10,254 people aged 45 and older from the Korean Longitudinal Study of Ageing from 2006 to 2016, 8,970 people were finalized as the study population. The dependent variable was whether a person has a normal cognitive function or not, and the independent variables of interest were BMI categorized by the World Health Organization Western Pacific Regional Office (WHO-WPRO) method, the WHO method, and a 10-group method. Covariates included sociodemographic factors, health behavior factors, and health status factors. A generalized linear mixed model analysis with a logit link was used. Results: In the adjusted model with all covariates, first, in the case of BMI categories of the WHO-WPRO method, underweight (odds ratio [OR], 1.16; 95% confidence interval [CI], 1.15-1.17), overweight (OR, 1.36; 95% CI, 1.35-1.36), and obese (OR, 1.34; 95% CI, 1.33-1.34) groups were more likely to have a normal cognitive function than a normal-weight group. Next, in the case of BMI categories of the WHO method, compared to a normal-weight group, underweight (OR, 1.15; 95% CI, 1.14-1.16) and overweight (OR, 1.06; 95% CI, 1.06-1.07) groups were more likely to have a normal cognitive function; however, obese (OR, 0.62; 95% CI, 0.61-0.63) group was less likely to have it. Lastly, in the case of the 10-group method, as BMI increased, the likelihood to have a normal cognitive function changed like a wave, reaching a global top at group-7 (26.5 kg/㎡ ≤ BMI <28.0 kg/㎡). Conclusion: The associations between BMI and cognitive function differed according to how BMI was categorized among people aged 45 and older in Korea, which suggests that cognitive function may be positively associated with BMI in some categories of BMI but negatively in its other categories. Health policies to reduce cognitive impairment need to consider this association between BMI and cognitive function.

• Journal of Korean Society of Forest Science
• /
• v.100 no.1
• /
• pp.70-78
• /
• 2011

In this study, we develop a fire stochastic simulation model by season based on the historical fire data in Korea. The model is utilized to generate sequences of fire events that are consistent with Korean fire history. We employ a three-stage approach. First, a random draw from a Bernoulli distribution is used to determine if any fire occurs for each day of a simulated fire season. Second, if a fire does occur, a random draw from a geometric multiplicity distribution determines their number. Last, ignition times for each fire are randomly drawn from a Poisson distribution. This specific distributional forms are chosen after analysis of Korean historical fire data. Maximum Likelihood Estimation (MLE) is used to estimate the primary parameters of the stochastic models. Fire sequences generated with the model appear to follow historical patterns with respect to diurnal distribution and total number of fires per year. We expect that the results of this study will assist a fire manager for planning fire suppression policies and suppression resource allocations.

• Journal of Korean Society of Forest Science
• /
• v.100 no.4
• /
• pp.540-547
• /
• 2011

This study was conducted to compare classification accuracy by tree species using QuickBird imagery for pixel- and segment-based classifications that have been mostly applied to classify land covers. A total of 398 points was used as training and reference data. Based on this points, the points were classified into fourteen land cover classes: four coniferous and seven deciduous tree species in forest classes, and three non-forested classes. In pixel-based classification, three images obtained by using raw spectral values, three tasseled indices, and three components from principal component analysis were produced. For the both classification processes, the maximum likelihood method was applied. In the pixel-based classification, it was resulted that the classification accuracy with raw spectral values was better than those by the other band combinations. As resulted that, the segment-based classification with a scale factor of 50% provided the most accurate classification (overall accuracy:76% and ${\hat{k}}$ value:0.74) compared to the other scale factors and pixel-based classification.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.41 no.6
• /
• pp.675-686
• /
• 2021

This study investigated the ground displacement occurring in a slope below a waste-rock dumping site and estimated the likelihood of a disaster due to a landslide. To start with, photogrammetry was conducted by unmanned aerial vehicles (UAVs) to investigate the size and extent of the ground displacement. From April 2019 to July 2020, the average error rate of the five UAV surveys was 0.011-0.034 m, and an elevation change of 2.97 m occurred due to the movement of the soil layer. Only some areas of the slope showedelevation change, and this was believed to be due to thegroundwater generated during rainfall rather than the effect of the waste-rock load at the top. Sensitivity analysis for LS-RAPID simulation was performed, and the simulation results were compared and analyzed by applying a digital elevation model (DEM) and a digital surface model (DSM)as terrain data with 10 m, 5 m, and 4 m grids. When data with high spatial resolution were used, the extent of the sedimentation of landslide material tended to be excessively expanded in the DEM. In contrast, in the result of applying a DSM, which reflects the topography in detail, the diffusion range was not significantly affected even when the spatial resolution was changed, and the sedimentation behavior according to the river shape could be accurately expressed. As a result, it was concluded that applying a DSM rather than a DEM does not significantly expand the sedimentation range, and results that reflect the site situation well can be obtained.

• Journal of Plant Biotechnology
• /
• v.49 no.3
• /
• pp.178-186
• /
• 2022

The tetraploid Solanum acaule is a wild potato species from Bolivia widely used for potato breeding because of its diverse attractive traits, including resistance to frost, late blight, potato virus X, potato virus Y, potato leafroll virus, potato spindle tuber viroid, and cyst nematode. However, the introgression of useful traits into cultivated potatoes via crossing has been limited by differences in endosperm balance number between species. Somatic fusion could be used to overcome sexual reproduction barriers and the development of molecular markers is essential to select proper fusion products. The chloroplast genome of S. acaule was sequenced using next-generation sequencing technology and specific markers for S. acaule were developed by comparing the obtained sequence with those of seven other Solanum species. The total length of the chloroplast genome is 155,570 bp, and 158 genes were annotated. Structure and gene content were very similar to other Solanum species and maximum likelihood phylogenetic analysis with 12 other species belonging to the Solanaceae family revealed that S. acaule is very closely related to other Solanum species. Sequence alignment with the chloroplast genome of seven other Solanum species revealed four InDels and 79 SNPs specific to S. acaule. Based on these InDel and SNP regions, one SCAR marker and one CAPS marker were developed to discriminate S. acaule from other Solanum species. These results will aid in exploring evolutionary aspects of Solanum species and accelerating potato breeding using S. acaule.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.22 no.4
• /
• pp.99-107
• /
• 2021

To establish an effective performance appraisal system, ratee accountability is essential because it increases the likelihood of meaningful utilization of performance appraisal feedback. In the context of a multi-rater performance appraisal system, this study examines the effect of the ratee general accountability and accountability to the supervisor and peers on the ratee's intention to change behavior based on the performance appraisal feedback. This study further explored the moderating effect of perceived feedback specificity on the relationship between the ratee accountability and behavioral change intention. Hierarchical multiple regression analysis was conducted using the survey data from 153 employees in eight firms with multi-rater performance appraisal systems. The results showed that ratee general accountability was positively associated with the intention to change behavior. Furthermore, perceived feedback specificity attenuated the positive relationship between the ratee general accountability and behavioral change intention, which was the opposite of the hypothesized direction. The findings indicate the importance of managing and facilitating ratee accountability for implementing a multi-rater performance appraisal system. The result also emphasizes the critical role that performance appraisal feedback plays in motivating performance improvement by providing flexible, constructive advice and behavioral change strategies instead of inducing self-defense or justification.

• Journal of Korea Water Resources Association
• /
• v.55 no.9
• /
• pp.657-668
• /
• 2022

With the occurrence of localized heavy rain while river flow has increased, both flow and rainfall cause riverside flood damages. As the degree of damage varies according to the level of social and economic impact, it is required to secure sufficient forecast lead time for flood response in areas with high population and asset density. In this study, the author established a flood risk matrix using ensemble rainfall runoff modeling and evaluated its applicability in order to increase the damage reduction effect by securing the time required for flood response. The flood risk matrix constructs the flood damage impact level (X-axis) using flood damage data and predicts the likelihood of flood occurrence (Y-axis) according to the result of ensemble rainfall runoff modeling using LENS rainfall data and as well as probabilistic forecasting. Therefore, the author introduced a method for determining the impact level of flood damage using historical flood damage data and quantitative flood damage assessment methods. It was compared with the existing flood warning data and the damage situation at the flood warning points in the Taehwa River Basin and the Hyeongsan River Basin in the Nakdong River Region. As a result, the analysis showed that it was possible to predict the time and degree of flood risk from up to three days in advance. Hence, it will be helpful for damage reduction activities by securing the lead time for flood response.

• Journal of the Korea Society for Simulation
• /
• v.18 no.3
• /
• pp.103-112
• /
• 2009

In a nuclear power plant (NPP), periodic sensor calibrations are required to assure sensors are operating correctly. However, only a few faulty sensors are found to be calibrated. For the safe operation of an NPP and the reduction of unnecessary calibration, on-line calibration monitoring is needed. In this paper, principal component-based Auto-Associative support vector regression (PCSVR) was proposed for the sensor signal validation of the NPP. It utilizes the attractive merits of principal component analysis (PCA) for extracting predominant feature vectors and AASVR because it easily represents complicated processes that are difficult to model with analytical and mechanistic models. With the use of real plant startup data from the Kori Nuclear Power Plant Unit 3, SVR hyperparameters were optimized by the response surface methodology (RSM). Moreover the statistical techniques are integrated with PCSVR for the failure detection. The residuals between the estimated signals and the measured signals are tested by the Shewhart Control Chart, Exponentially Weighted Moving Average (EWMA), Cumulative Sum (CUSUM) and generalized likelihood ratio test (GLRT) to detect whether the sensors are failed or not. This study shows the GLRT can be a candidate for the detection of sensor drift.

• Journal of Korean Neurosurgical Society
• /
• v.66 no.5
• /
• pp.536-542
• /
• 2023

Objective : Surgical site infection is the most detrimental complication following cranioplasty. In other surgical fields, intrawound vancomycin powder application has been introduced to prevent surgical site infection and is widely used based on results in multiple studies. This study evaluated the effect of intrawound vancomycin powder in cranioplasty compared with the conventional method without topical antibiotics. Methods : This retrospective study included 580 patients with skull defects who underwent cranioplasty between August 1, 1998 and December 31, 2021. The conventional method was used in 475 (81.9%; conventional group) and vancomycin powder (1 g) was applied on the dura mater and bone flap in 105 patients (18.1%; vancomycin powder group). Surgical site infection was defined as infection of the incision, organ, or space that occurred after cranioplasty. Surgical site infection within 1-year surveillance period was compared between the conventional and vancomycin powder groups with logistic regression analysis. Penalized likelihood estimation method was used in logistic regression to deal with zero events. All local and systemic adverse events associated with topical vancomycin application were also evaluated. Results : Surgical site infection occurred in 31 patients (5.3%) and all were observed in the conventional group. The median time between cranioplasty and detection of surgical site infection was 13 days (range, 4-333). Staphylococci were the most common organisms and identified in 25 (80.6%) of 31 cases with surgical site infections. The surgical site infection rate in the vancomycin powder group (0/105, 0.0%) was significantly lower than that in the conventional group (31/475, 6.5%; crude odds ratio [OR], 0.067; 95% confidence interval [CI], 0.006-0.762; adjusted OR, 0.068; 95% CI, 0.006-0.731; p=0.026). No adverse events associated with intrawound vancomycin powder were observed during the follow-up. Conclusion : Intrawound vancomycin powder effectively prevented surgical site infections following cranioplasty without local or systemic adverse events. Our results suggest that intrawound vancomycin powder is an effective and safe strategy for patients undergoing cranioplasty.