• Journal of Korean Society of Environmental Engineers
• /
• v.37 no.3
• /
• pp.137-144
• /
• 2015

The purpose of this study is to develop an warning system to detect real-time water quality abnormality using a multivariate statistical approach. In this study, we applied principal component analysis among multivariate data analyses which was used for the correlation between water quality parameters considering the real-time algorithm to determine abnormality in water quality. We applied our approach to real field data and showed the utilization of algorithm for the real-time monitoring to find water quality abnormality. In addition, our approach with Korea Meterological Adminstration database identified heavy rain data due to climate change is one of the most important factors to explain water quality abnormality.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.46 no.4
• /
• pp.1-12
• /
• 2009

An RFID application needs to collect product's information scattered over the RFID network efficiently according to the globalization of RFID applied enterprises. To be informed of the stock status of products promptly in the supply chain network, especially, it is necessary to support queries that retrieve statistical information of tagged products. Since existing RFID network does not provide these kinds of queries, however, an application should request a query to several RFID middleware systems and analyze collected data directly. This process makes an application do a heavy computation for retrieving statistical information. To solve this problem, we define a new Distributed Continuous Query that finds information of tagged products from the global RFID network and provides statistical information to RFID applications. We also propose a Distributed Continuous Query System to process the distributed continuous query efficiently. To find out the movement of products via multiple RFID systems in real time, our proposed system uses Pedigree which represents trade information of items. Our system can also reduce the cost of query processing for removing duplicated data from multiple middleware systems by using Pedigree.

• Smart Media Journal
• /
• v.9 no.3
• /
• pp.80-89
• /
• 2020

Out - of - distribution (OOD) samples are frequently encountered when deploying a classification model in plenty of real-world machine learning-based applications. Those samples are normally sampling far away from the training distribution, but many classifiers still assign them high reliability to belong to one of the training categories. In this study, we address the problem of removing OOD examples by estimating marginal density estimation using variational autoencoder (VAE). We also investigate other proper methods, such as temperature scaling, Gaussian discrimination analysis, and label smoothing. We use Chonnam National University (CNU) weeds dataset as the in - distribution dataset and CIFAR-10, CalTeach as the OOD datasets. Quantitative results show that the proposed framework can reject the OOD test samples with a suitable threshold.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.29 no.8 s.145
• /
• pp.1102-1113
• /
• 2005

The propose of this study is to develop the optimal sizing system of ready-to-wear f3r elementary school girls using a newly invented statistical technique. The body measurements was classified by the method that equalizes the distribution of the subjects using the probability density function, to theoretically systemize a method to determine a size range of ready-to-wear for elementary school girls between 6 to 12 years old. The statistical method were 1) The total of 11 height groups, which size interval from one another is 6 cm that is an average height gap between each age. 2) In order to determine an approximate figure (m ${\times}$ n) to establish the appropriate sizes far each height group that fit to the combinations of bust and hip girth, which based on their means and standard deviations on the probability density curve to produce the standard normal distribution. 3) m and n were aligned by 4cm -the grading increments used for patterns making- and determined the size ranges by confirming the approximate figures of m and n. 4) The representative values were determined by an area ratio calculated by dividing the area determined from the range of bust and hip girth with the representative value. Considering the characteristics of subjects' distribution, the area ratios was used. 5) Weight was calculated by seeking a growth exponent for each age and multiplying it by the number of girls that fit to each size range. As sections that show the highest weight are more likely sought by the consumers, these sections were determined as the optimal size standards. 6) This optimal sizing system consists of sizes determined by the optimal size standards and its sizes are marked with height, bust and hip girth.

• Atmosphere
• /
• v.30 no.4
• /
• pp.421-437
• /
• 2020

The effects of the terminal fall velocity-diameter relationship for raindrops, which is prescribed based on the measurement, on the simulated surface precipitation over Korea during summer season were investigated in our study. Two rainfall cases, 1-month summer precipitation and mesoscale rainfall, have been simulated using the Weather Research and Forecasting (WRF) model. The selected cloud microphysics parameterizations are WRF Single-Moment 5-class (WSM5) and WRF Single-Moment 6-class (WSM6) in the WRF model. The measured terminal fall-diameter relationship for raindrops by Gunn and Kinzer (1949) was applied in both WSM5 and WSM6. The sensitivity experiments with WSM5 and WSM6, applying the measured fall-diameter relationship, presents the different responses in simulated precipitation amount for the 1-month summer precipitation case. Precipitation increases with WSM5, thus enhancing the precipitation statistical skills. However, precipitation decreases with WSM6 leading to the deterioration of precipitation statistical skills. For the mesoscale rainfall case, precipitation increases with both WSM5 and WSM6, which further enhances the positive bias in precipitation amount.

• Journal of Korean Society for Atmospheric Environment
• /
• v.24 no.E2
• /
• pp.92-101
• /
• 2008

The objective of this study is to evaluate and improve the capability of the Weather Research and Forecasting (WRF) model in simulating fog and visibility in local airports over Korea. The WRF model system is statistically evaluated for the 48-fog cases over Korea from 2003 to 2006. Based on the 4-yr evaluations, attempts are made to improve the simulation skill of fog and visibility over Korea by revising the statistical coefficients in the visibility algorithms of the WRF model. A comparison of four existing visibility algorithms in the WRF model shows that uncertainties in the visibility algorithms include additional degree of freedom in accuracy of numerical fog forecasts over Korea. A revised statistical algorithm using a linear-regression between the observed visibility and simulated hydrometeors and humidity near the surface exhibits overall improvement in the visibility forecasts.

• Communications for Statistical Applications and Methods
• /
• v.18 no.4
• /
• pp.495-505
• /
• 2011

This study performs statistical tests for stability of a long-run relationship in the telecommunication market system by identifying the time path of a recursively estimated cointegration parameter. A dummy variable is used to recover stability for the period that the hypothesis of stable cointegration is rejected, and then a proper cointegrating relation is derived. A dummy variable appears to reflect the structural change in the cointegrating relation according to the analytical results for the error correction term.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.38 no.3
• /
• pp.87-94
• /
• 2015

Recently, management based on statistical data has become a big issue and the importance of the statistics has been emphasized for the management innovation in the defense area. However, the Military Management based on the statistics is hard to expect because of the shortage of the statistics in the military. There are many military information systems having great many data created in real time. Since the infrastructure for gathering data form the many systems and making statistics by using gathered data is not equipped, the usage of the statistics is poor in the military. The Analytical Defense Statistics System is designed to improve effectively the defense management in this study. The new system having the sub-systems of Data Management, Analysis and Service can gather the operational data from interlocked other Defense Operational Systems and produce Defense Statistics by using the gathered data beside providing statistics services. Additionally, the special function for the user oriented statistics production is added to make new statistics by handling many statistics and data. The Data Warehouse is considered to manage the data and Online Analytical Processing tool is used to enhance the efficiency of the data handling. The main functions of the R, which is a well-known analysis program, are considered for the statistical analysis. The Quality Management Technique is applied to find the fault from the data of the regular and irregular type. The new Statistics System will be the essence of the new technology like as Data Warehouse, Business Intelligence, Data Standardization and Statistics Analysis and will be helpful to improve the efficiency of the Military Management.

• Journal of Korea Multimedia Society
• /
• v.21 no.8
• /
• pp.837-848
• /
• 2018

Monitoring technology of machining has a long history since unmanned machining was introduced. Despite the long history, many researchers have presented new approaches continuously in this area. Sound based machine fault diagnosis is the process consisting of detecting automatically the damages that affect the machines by analyzing the sounds they produce during their operating time. The collected sound is corrupted by the surrounding work environment. Therefore, the most important part of the diagnosis is to find hidden elements inside the data that can represent the error pattern. This paper presents a feature extraction methodology that combines various digital signal processing and pattern recognition methods for the analysis of the sounds produced by tools. The magnitude spectrum of the sound is extracted using the Fourier analysis and the band-pass filter is applied to further characterize the data. Statistical functions are also used as input to the nonlinear classifier for the final response. The results prove that the proposed feature extraction method accurately captures the hidden patterns of the sound generated by the tool, unlike the conventional features. Therefore, it is shown that the proposed method can be applied to a sound based automatic diagnosis system.

• Journal of information and communication convergence engineering
• /
• v.9 no.4
• /
• pp.475-482
• /
• 2011

In this research, we have attempted to estimate the water quality of fish farms in terms of parameters such as water temperature, dissolved oxygen, pH, and salinity by employing observational data obtained from a coastal ocean observatory of a national institution located close to the fish farm. We requested and received marine data comprising nine factors including water temperature from Korea Hydrographic and Oceanographic Administration. For verifying our results, we also established an experimental fish farm in which we directly placed the sensor module of an optical mode, YSI-6920V2, used for self-cleaning inside fish tanks and used the data measured and recorded by a environment monitoring system that was communicating serially with the sensor module. We investigated the differences in water temperature and salinity among three areas - Goheung Balpo, Yeosu Odongdo, and the experimental fish farm, Keumho. Water temperature did not exhibit significant differences but there was a difference in salinity (significance <5%). Further, multiple regression analysis was performed to estimate the water quality of the fish farm at Keumho based on the data of Goheung Balpo. The water temperature and dissolved-oxygen estimations had multiple regression linear relationships with coefficients of determination of 98% and 89%, respectively. However, in the case of the pH and salinity estimated using the oceanic environment with nine factors, the adjusted coefficient of determination was very low at less than 10%, and it was therefore difficult to predict the values. We plotted the predicted and measured values by employing the estimated regression equation and found them to fit very well; the values were close to the regression line. We have demonstrated that if statistical model equations that fit well are used, the expense of fish-farm sensor and system installations, maintenances, and repairs, which is a major issue with existing environmental information monitoring systems of marine farming areas, can be reduced, thereby making it easier for fish farmers to monitor aquaculture and mariculture environments.