• International Journal of Contents
• /
• v.7 no.3
• /
• pp.1-5
• /
• 2011

It has been an interesting challenge to find a good classifier for imbalanced data, since it is pervasive but a difficult problem to solve. However, classifiers developed with the assumption of well-balanced class distributions show poor classification performance for the imbalanced data. Among many approaches to the imbalanced data problems, the algorithmic level approach is attractive because it can be applied to the other approaches such as data level or ensemble approaches. Especially, the error back-propagation algorithm using the target node method, which can change the amount of weight-updating with regards to the target node of each class, attains good performances in the imbalanced data problems. In this paper, we analyze the relationship between two optimal outputs of neural network classifier trained with the target node method. Also, the optimal relationship is compared with those of the other error function methods such as mean-squared error and the n-th order extension of cross-entropy error. The analyses are verified through simulations on a thyroid data set.

• Proceedings of the IEEK Conference
• /
• 1999.11a
• /
• pp.487-490
• /
• 1999

In this paper we propose an efficient content-based image retrieval method using the color and wavelet based features. The color features are extracted from color histograms of the global image and the wavelet based features are extracted from the invariant moments of the high-pass band image through the spatial-frequency analysis of the wavelet transform. The proposed algorithm, called color and wavelet features based query(CWBQ), is composed of two-step query operations for efficient image retrieval: the coarse level filtering operation and the fine level matching operation. In the first filtering operation, the color histogram feature is used to filter out the dissimilar images quickly from a large image database. The second matching operation applies the wavelet based feature to the retained set of images to retrieve all relevant images successfully. The experimental results show that the proposed algorithm yields more improved retrieval accuracy with computationally efficiency than the previous methods.

• Korean Journal of Health Education and Promotion
• /
• v.32 no.5
• /
• pp.53-62
• /
• 2015

Objectives: The purpose of this study was to examine the associations among self-rated health and socioeconomic status. Methods: Analyses were conducted based on cross-sectional data obtained from the Korea Youth Risk Behavior Web-based Survey. A total of 79,202 students aged 12 to 18 years participated in the study and there was a response rate of 95.5%. Separate logistic regression analyses were performed on each gender group based on a set of independent variables. Those being: the level of parental education level; family affluence scale; subjective household economic status; and subjective school achievement with SRH as the dependent variable. Results: Multivariate analyses revealed significant associations between each SES and adolescent SRH after controlling for other covariates. However, in the models that included all SES indicators, subjective household economic status and subjective school achievement remained significant in boys and girls. Conclusions: The findings demonstrated that subjective SES indicators are more closely related to adolescent SRH when compared with objective indicators.

• Management & Information Systems Review
• /
• v.15
• /
• pp.147-164
• /
• 2004

The purpose of this paper is on employees as internal customers and the critical role this group plays in the delivery of quality results. The set up of research model for verification was as follows. The research model was drawn as internal service quality level $\Rightarrow$ internal customer satisfaction $\Rightarrow$ enterprise outcome. Then, two hypotheses were established to the research model. Through the factor analysis and multiple regression analysis, the results are as follows. First, internal service quality level turned out to be affected indirectly through internal customers' satisfaction rather than a direct factor to affect the enterprise outcome. Second, internal customers' satisfaction was proved to be the most important factor for the enterprise outcome as ti was the intimate factor precedent to the enterprise outcome. However, there could be a variation of response according to the personal circumstances of respondents since the respondents were from different enterprises and consisted various job positions and age group. Namely it included a limitation of rather unaccurate resulting values because the transverse methods were performed for convenience though it needed a longitudinal research to accomplish the general purpose of this study.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.11
• /
• pp.3493-3506
• /
• 2022

In the field of scene segmentation, the precise segmentation of object boundaries in sports movement scene images is a great challenge. The geometric information and spatial information of the image are very important, but in many models, they are usually easy to be lost, which has a big influence on the performance of the model. To alleviate this problem, a parallel dense dilated convolution merging Network (termed PDDCM-Net) was proposed. The proposed PDDCMNet consists of a feature extractor, parallel dilated convolutions, and dense dilated convolutions merged with different dilation rates. We utilize different combinations of dilated convolutions that expand the receptive field of the model with fewer parameters than other advanced methods. Importantly, PDDCM-Net fuses both low-level and high-level information, in effect alleviating the problem of accurately segmenting the edge of the object and positioning the object position accurately. Experimental results validate that the proposed PDDCM-Net achieves a great improvement compared to several representative models on the COCO-Stuff data set.

• International Journal of Computer Science & Network Security
• /
• v.23 no.12
• /
• pp.101-106
• /
• 2023

Chronic illnesses are among the most common serious problems affecting human health. Early diagnosis of chronic diseases can assist to avoid or mitigate their consequences, potentially decreasing mortality rates. Using machine learning algorithms to identify risk factors is an exciting strategy. The issue with existing feature selection approaches is that each method provides a distinct set of properties that affect model correctness, and present methods cannot perform well on huge multidimensional datasets. We would like to introduce a novel model that contains a feature selection approach that selects optimal characteristics from big multidimensional data sets to provide reliable predictions of chronic illnesses without sacrificing data uniqueness.[1] To ensure the success of our proposed model, we employed balanced classes by employing hybrid balanced class sampling methods on the original dataset, as well as methods for data pre-processing and data transformation, to provide credible data for the training model. We ran and assessed our model on datasets with binary and multivalued classifications. We have used multiple datasets (Parkinson, arrythmia, breast cancer, kidney, diabetes). Suitable features are selected by using the Hybrid feature model consists of Lassocv, decision tree, random forest, gradient boosting,Adaboost, stochastic gradient descent and done voting of attributes which are common output from these methods.Accuracy of original dataset before applying framework is recorded and evaluated against reduced data set of attributes accuracy. The results are shown separately to provide comparisons. Based on the result analysis, we can conclude that our proposed model produced the highest accuracy on multi valued class datasets than on binary class attributes.[1]

• International Journal of Highway Engineering
• /
• v.17 no.5
• /
• pp.57-66
• /
• 2015

PURPOSES: Nighttime driving is very different from daytime driving because drivers must obtain nighttime sight-distances based on road lights and headlights. Unfortunately, nighttime driving conditions in Korea are far from ideal due to poor lighting and an insufficient number of road lights and inadequate operation and maintenance of delineators. This study is conducted to develop new standards for nighttime road visibility based on experiments of driver perception for nighttime visibility conditions. METHODS : In the study, perception level and satisfaction of nighttime visibility were investigated. A total of 60 drivers participated, including 34 older drivers and 31 young drivers. To evaluate driver perceptions of nighttime road visibility, fuzzy set theory was used because the conventional analysis methods for driver perception are limited in effectiveness for considering the characteristics of perception which are subjective and vague, and are generally expressed in terms of linguistic terminologies rather than numerical parameters. RESULTS : This study found that levels of nighttime visibility, as perceived by drivers, are remarkably similar to their satisfactions in different nighttime driving conditions with a log-function relationship. Older drivers evaluated unambiguously degree of nighttime visibility but evaluations by young drivers regarding it were unclear. CONCLUSIONS : A minimum value of brightness on roads was established as YUX 30, based on final analyzed results. In other words, road lights should be installed and operated to obtain more than YUX 30 brightness for the safety and comfort of nighttime driving.

• Journal of Korean society of Dental Hygiene
• /
• v.13 no.4
• /
• pp.589-596
• /
• 2013

Objectives : The aim of the study is to investigate the influence of the stress of adults on their oral mucosal diseases, dry mouth and physical, mental stress symptoms. Structured equation model (SEM) was used to analyze the hypotheses of the study. Methods : The subjects were 500 adultsfrom July 1 to December 31, 2012. The data were analyzed using SPSS 18.0 (SPSS 18.0 K for window, SPSS Inc USA) and IBM SPSS Amos 18.0 (SPSS Inc, Chicago, IL, USA) set at the level of significance as 0.05. Results : The level of stress had a direct influence on oral mucosal diseases, and oral mucosal diseases affected stress symptoms directly. The level of stress had a significant impact on stress symptoms, and that exercised an indirect influence on stress symptoms through the medium of oral mucosal diseases and dry mouth. The level of stress affected dry mouth in a direct effects, and dry mouth had a direct impact on stress symptoms. Conclusions : The stress of adults had direct and indirect impacts on their oral health and systemic diseases. The oral health of adults should be promoted to let them stay healthy, and how to help them to get rid of their stress should be considered to improve their quality of life.

• Analytical Science and Technology
• /
• v.8 no.4
• /
• pp.1079-1087
• /
• 1995

The quality of human life is directly related to the quality of the environment. To assess environmental quality we must first determine the MCLG(Maximum Contaminant Level Goal), MCL(Maximum Contaminant Level), environmental impact and so on. The MCLG is the concentration at which no known adverse health effects occur. The MCLG is determined by risk assessment identifying which process is hazardous assessing, dose-response, human exposure, and characteristics of risk. With consideration of analytical methods, treatment technology, cost and regulatory impact, the MCL is set as close to the MCLG as possible. In this way, determination of the concentration and national distribution of contaminants is important for assessment of environmental quality The analytical sciences pose potential problems in assessing environmental quality. Continuing improvement in the performance of analytical instruments and operating technique has been lowering the limits of detectability. Contaminant concentration below the detection limit has usually been reported as ND(Not-Detected) and this has often been misunderstood as equivalent to zero. Because of this, more the contaminant concentration in the past was below the detection limit, whereas contaminants can be quantified now even though the contaminant concentration might remain the same or may even have decreased. In addition, environmental sampling has various components due to heterogeneous matrices. These samples are used to overestimate the concentration of the contaminant due to large variability, resulting in excess readings for MCL. In this paper, the significance of the analytical sciences is emphasized in both a conceptual and a technical approach to environmental assessment.

• Journal of Korean Academy of Fundamentals of Nursing
• /
• v.24 no.1
• /
• pp.60-71
• /
• 2017

Purpose: The purpose of this study was to develop video assisted education on bowel preparation for colonoscopy (VEBPC) and use a snartphone to evaluate effects of the VEBPC. Methods: Adult patients who were scheduled for colonoscopy were recruited from a university general hospital and randomly assigned to three groups. Group 1 (n=30) watched the video using a computer set in the endoscope consulting room. Group 2 (n=29) watched it using a smartphone, and group 3, the control group (n=29) received education with existing instructions at the reservation-reception desk. Participants were evaluated on knowledge on taking bowel preparation agents and diet, compliance on taking bowel preparation agents and diet, satisfaction with education, and actual level of bowel preparation. Results: Group 1 and 2 showed significantly (p<.001) higher scores for knowledge, compliance, and satisfaction compared to the control group. However, in post-hoc test analyses there were no significant differences in these variables between group 1 and 2. No significant difference was found in the actual level of bowel preparation among the three groups. Conclusion: Findings from this study show that VEBPC using smartphone is a better option than existing educational methods. However, replication studies are necessary to confirm these findings.