• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.57-73
• /
• 2021

Maintenance and prevention of failure through anomaly detection of ICT infrastructure is becoming important. System monitoring data is multidimensional time series data. When we deal with multidimensional time series data, we have difficulty in considering both characteristics of multidimensional data and characteristics of time series data. When dealing with multidimensional data, correlation between variables should be considered. Existing methods such as probability and linear base, distance base, etc. are degraded due to limitations called the curse of dimensions. In addition, time series data is preprocessed by applying sliding window technique and time series decomposition for self-correlation analysis. These techniques are the cause of increasing the dimension of data, so it is necessary to supplement them. The anomaly detection field is an old research field, and statistical methods and regression analysis were used in the early days. Currently, there are active studies to apply machine learning and artificial neural network technology to this field. Statistically based methods are difficult to apply when data is non-homogeneous, and do not detect local outliers well. The regression analysis method compares the predictive value and the actual value after learning the regression formula based on the parametric statistics and it detects abnormality. Anomaly detection using regression analysis has the disadvantage that the performance is lowered when the model is not solid and the noise or outliers of the data are included. There is a restriction that learning data with noise or outliers should be used. The autoencoder using artificial neural networks is learned to output as similar as possible to input data. It has many advantages compared to existing probability and linear model, cluster analysis, and map learning. It can be applied to data that does not satisfy probability distribution or linear assumption. In addition, it is possible to learn non-mapping without label data for teaching. However, there is a limitation of local outlier identification of multidimensional data in anomaly detection, and there is a problem that the dimension of data is greatly increased due to the characteristics of time series data. In this study, we propose a CMAE (Conditional Multimodal Autoencoder) that enhances the performance of anomaly detection by considering local outliers and time series characteristics. First, we applied Multimodal Autoencoder (MAE) to improve the limitations of local outlier identification of multidimensional data. Multimodals are commonly used to learn different types of inputs, such as voice and image. The different modal shares the bottleneck effect of Autoencoder and it learns correlation. In addition, CAE (Conditional Autoencoder) was used to learn the characteristics of time series data effectively without increasing the dimension of data. In general, conditional input mainly uses category variables, but in this study, time was used as a condition to learn periodicity. The CMAE model proposed in this paper was verified by comparing with the Unimodal Autoencoder (UAE) and Multi-modal Autoencoder (MAE). The restoration performance of Autoencoder for 41 variables was confirmed in the proposed model and the comparison model. The restoration performance is different by variables, and the restoration is normally well operated because the loss value is small for Memory, Disk, and Network modals in all three Autoencoder models. The process modal did not show a significant difference in all three models, and the CPU modal showed excellent performance in CMAE. ROC curve was prepared for the evaluation of anomaly detection performance in the proposed model and the comparison model, and AUC, accuracy, precision, recall, and F1-score were compared. In all indicators, the performance was shown in the order of CMAE, MAE, and AE. Especially, the reproduction rate was 0.9828 for CMAE, which can be confirmed to detect almost most of the abnormalities. The accuracy of the model was also improved and 87.12%, and the F1-score was 0.8883, which is considered to be suitable for anomaly detection. In practical aspect, the proposed model has an additional advantage in addition to performance improvement. The use of techniques such as time series decomposition and sliding windows has the disadvantage of managing unnecessary procedures; and their dimensional increase can cause a decrease in the computational speed in inference.The proposed model has characteristics that are easy to apply to practical tasks such as inference speed and model management.

• Journal of Bio-Environment Control
• /
• v.31 no.4
• /
• pp.384-392
• /
• 2022

The objective of this study was to estimated nitrogen content and chlorophyll using RGB, Hyperspectral sensors to diagnose of nitrogen nutrition in apple tree leaves. Spectral data were acquired through image processing after shooting with high resolution RGB and hyperspectral sensor for two-year-old 'Hongro/M.9' apple. Growth data measured chlorophyll and leaf nitrogen content (LNC) immediately after shooting. The growth model was developed by using regression analysis (simple, multi, partial least squared) with growth data (chlorophyll, LNC) and spectral data (SPAD meter, color vegetation index, wavelength). As a result, chlorophyll and LNC showed a statistically significant difference according to nitrogen fertilizer level regardless of date. Leaf color became pale as the nutrients in the leaf were transferred to the fruit as over time. RGB sensor showed a statistically significant difference at the red wavelength regardless of the date. Also hyperspectral sensor showed a spectral difference depend on nitrogen fertilizer level for non-visible wavelength than visible wavelength at June 10th and July 14th. The estimation model performance of chlorophyll, LNC showed Partial least squared regression using hyperspectral data better than Simple and multiple linear regression using RGB data (Chlorophyll R²: 81%, LNC: 81%). The reason is that hyperspectral sensor has a narrow Full Half at Width Maximum (FWHM) and broad wavelength range (400-1,000 nm), so it is thought that the spectral analysis of crop was possible due to stress cause by nitrogen deficiency. In future study, it is thought that it will contribute to development of high quality and stable fruit production technology by diagnosis model of physiology and pest for all growth stage of tree using hyperspectral imagery.

• Research in Community and Public Health Nursing
• /
• v.14 no.2
• /
• pp.287-298
• /
• 2003

The purpose of the study was to develop an integrated prevention program to strengthen elders self-care capability and to examine its effectiveness on their psychological condition. This study used one group pre- and post-test design. Subjects were 85 elderly residents (over 65 years of age) who lived alone, and received free basic medical care and social welfare services in a rural community in Korea. Subject eligibility criteria for this study were to an elders who 1) is not currently taking any anti-depressant medication 2) is able to communicate, and 3) agrees to participate in this study. The integrated program was composed of horticulture, reminiscence, and friendship activities. Twelve sessions were provided for 12 weeks in community-based partnerships to achieve better outcomes. The intervention was case-managed by a public health nurse and aided by six volunteers. The main outcome variable was depression, which was assessed by using 15 items selected from the Geriatric Depression Scale-short form Korean version. Socio-demographic characteristics, functional status, and satisfaction with social support were used as covariates. Results showed that there was a significant intervention effect at post-intervention time point compared to pre-intervention time point(E.S. 0.94). Multiple linear regression analysis showed significant interaction effects between intervention and satisfaction with social support. These findings must be interpreted within the context that an effects of an integrated program could be more synergistically increased when social support factor is considered in the program. A community-based integrated prevention program of depression is effective for vulnerable rural elderly. It is suggested that randomized controlled trials within community setting for better methodological strength as well as multi-level outcomes on community need to be conducted in future.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.10 no.2
• /
• pp.111-122
• /
• 2010

As the Jamsu-bridge and the floodplains of the Han River can be flooded during the rainy season, the exact prediction of the peak flood time is very important for mitigation of flood hazard. This study analyzes the effect of outflow of Paldang Dam and tide of Yellow Sea on the Han River. A target area is from the Paldang dam to Jeonryu gauging station. Water level of Jeonryu as a downstream boundary condition was estimated through multi linear regression analysis with outflow of Paldang dam and tide level of Incheon, because it was influenced by both a tide of Yellow Sea and outflow of Paldang dam. In this study, Water Level Rising Travel Time of the Jamsu-bridge and some floodplains in the Han River are estimated. Also, The second order polynomial expressions for relationships of outflow of Paldang Dam and Water Level Rising Travel Time were developed considering the outflow of Paldang dam and tide of Yellow Sea.

• Korean Security Journal
• /
• no.44
• /
• pp.37-58
• /
• 2015

The purpose of this study was to verify the effect of passion on major students for security martial arts. Participants were selected 393 undergraduate students who study security martial arts. Sampling was used by purposive sample method. Measurements were used with multi-dimensional passion and motivation scales. Statistic was executed descriptive statistics, exploratory factor analysis, Pearson's correlation and multiple linear regression. The results of this study were as follows: First, motivation of student for security martial arts has influencing relationship with harmonious passion. Second, motivation has influencing relationship with obsessive passion. Third, motivation has influencing relationship with sense of unity. Forth, motivation has influencing relationship with preference of passion. Fifth, motivation has influencing relationship with value of passion. Sixth, motivation has influencing relationship with investment of passion. That is, internal motivation such as giving value and inner satisfaction gives positive effect to passion above mentioned factors but a motivation gives negative effect to passion.

• Journal of Korean Academy of Nursing
• /
• v.41 no.6
• /
• pp.780-787
• /
• 2011

Purpose: This study was conducted to examine activities of daily living (ADL) of older adults admitted to Korean long-term care hospitals (LTCHs), and to explore the patient and organizational factors that have an impact on the ADL of this population. Methods: A secondary analysis of the Korean minimum data set (K-MDS) of patients (N=14,369) and of the profiles of LTCHs (N=358) from the Health Insurance Review and Assessment Service was done between January and July 2008. The outcome variable was ADL score 6 months after baseline assessment. Multi-level linear regression was employed to explore the patient and organizational factors that affected ADL scores. Results: Of the patients, 45.4% had a baseline ADL score of between 31 and 40, with a score of 40 indicating that the patient was entirely dependent for all items. None of the organizational characteristics were significantly associated with effects on the ADLs of older adults who had been in a LTHC for at least 6 months. However, patient characteristics, such as age, baseline ADL, frequency of physical therapy, urinary incontinence, fecal incontinence, pressure ulcers, and having a tube or catheter, were significantly associated with ADL 6 months after baseline. Conclusion: In order to maintain and improve the ADL of older adults in LTCHs, we should develop strategies to prevent urinary and fecal incontinence, pressure ulcers, unnecessary tubes or catheters, providing adequate physical therapy. Additional studies should include more detailed information regarding nursing staff, including RN hours for direct care, education level and turnover.

• Nutrition Research and Practice
• /
• v.6 no.5
• /
• pp.458-465
• /
• 2012

This study aims to determine the association between obesity-related behaviors (dietary practices, physical activity and body image) and body weight status among adolescents. A total of 382 adolescents (187 males and 195 females) aged 13 to 15 years in Kajang, Selangor participated in this study. Majority of the respondents were Malays (56.0%), followed by Chinese (30.1%) and Indians (13.9%). Dietary practices, physical activity and body image of the adolescents were assessed through the eating behaviors questionnaire, two-day dietary record, two-day physical activity record and multi-dimensional body image scale (MBIS), respectively. Body weight and height were measured by trained researchers. The prevalence of overweight and obesity (19.5%) was about twice the prevalence of underweight (10.5%). About two-thirds of the respondents (72.3%) skipped at least one meal and half of them (56.2%) snacked between meals with a mean energy intake of $1,641{\pm}452$ kcal/day. More than half of the respondents (56.8%) were practicing sedentary lifestyle with a mean energy expenditure of $1,631{\pm}573$ kcal per day. Energy intake (r = 0.153, P < 0.05), physical activity (r = 0.463, P < 0.01) and body image (r = 0.424, P < 0.01) were correlated with BMI. However, meal skipping, snacking and energy expenditure per kg body weight were not associated with body weight status. Multiple linear regression analysis showed that body image, physical activity and energy intake contributed significantly in explaining body weight status of the adolescents. In short, overweight and obesity were likely to be associated not only with energy intake and physical activity, but also body image. Hence, promoting healthy eating, active lifestyle and positive body image should be incorporated in future obesity prevention programmes in adolescents.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.30 no.6_1
• /
• pp.529-538
• /
• 2012

By using an algorithm derived by a multiple linear regression analysis, a technique for filtering was developed; and by using the developed technique, the results of conducting filtering of the raw data collected via scanning with a terrestrial LiDAR the actual sloped terrain was analyzed. As such, when filtering was applied by dividing the observation areas into two areas with the topographical line as a reference in order to improve the filtering accuracy, it was seen that the filtering accuracy improved by about 8.73% as compared to when filtering was applied without dividing the observation area. In addition, considering the fact that the accuracy improved by 5~7% when the sloped sides of a multicurvature topography were divided and a complex filtering applied as compared to when filtering was applied for the entire area or by regions, it can be asserted that the accuracy was higher when a complex filtering was conducted by dividing the sloped areas where the slope is not constant due to the multi-curvature of topography.

• Journal of Korea Society of Industrial Information Systems
• /
• v.28 no.6
• /
• pp.107-113
• /
• 2023

In the field of ammunition, storage performance is recognized as important, and the moisture-proof performance of packaging containers is very important. In the ammunition field, paper cans with a multi-layered structure are mainly used as packaging containers. It is made by layering materials that play various roles. These packaging containers are mainly evaluated for moisture-proof performance according to the Korean Industrial Standard KS T 1314. The moisture permeability is determined through linear regression analysis of the change in weight of the moisture absorbent added inside. In this study, the effect of the amount of desiccant added on the moisture permeability test results of packaging containers was confirmed. It is considered appropriate that the amount of desiccant used in testing ammunition packaging containers be approximately 70% or more of the internal volume.

• The Journal of Engineering Geology
• /
• v.18 no.3
• /
• pp.263-276
• /
• 2008

In diverse hydrogeologic fields, estimation of groundwater storage change is one of the most critical issues. Accurate estimation methods for determining groundwater storage change are required more and more. For Yeonyang area of Ulsan Megacity, groundwater storage change was estimated by using water balance method and hydrogeological analyses. The estimates of groundwater storage change was 240 mm corresponding to 18.7% of mean annual precipitation. Direct runoff was calculated as 137 mm (10.6% of mean annual precipitation) by using SCS-CN method. Evapotranspiration based on the Thornthwaite method was calculated as 776 mm (60.5% of mean annual precipitation). Hydraulic properties of the soil types do not show any distinct relation with hydraulic conductivity of the rocks. This fact suggests that hydraulic property on the surface is different from that of subsurface geology. According to multi-linear regression analysis between groundwater storage change and hydraulic parameters, a regression equation of groundwater storage change, which was explained by precipitation and evapotranspiration, was established.