• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.23-45
• /
• 2020

Big data is creating in a wide variety of fields such as medical care, manufacturing, logistics, sales site, SNS, and the dataset characteristics are also diverse. In order to secure the competitiveness of companies, it is necessary to improve decision-making capacity using a classification algorithm. However, most of them do not have sufficient knowledge on what kind of classification algorithm is appropriate for a specific problem area. In other words, determining which classification algorithm is appropriate depending on the characteristics of the dataset was has been a task that required expertise and effort. This is because the relationship between the characteristics of datasets (called meta-features) and the performance of classification algorithms has not been fully understood. Moreover, there has been little research on meta-features reflecting the characteristics of multi-class. Therefore, the purpose of this study is to empirically analyze whether meta-features of multi-class datasets have a significant effect on the performance of classification algorithms. In this study, meta-features of multi-class datasets were identified into two factors, (the data structure and the data complexity,) and seven representative meta-features were selected. Among those, we included the Herfindahl-Hirschman Index (HHI), originally a market concentration measurement index, in the meta-features to replace IR(Imbalanced Ratio). Also, we developed a new index called Reverse ReLU Silhouette Score into the meta-feature set. Among the UCI Machine Learning Repository data, six representative datasets (Balance Scale, PageBlocks, Car Evaluation, User Knowledge-Modeling, Wine Quality(red), Contraceptive Method Choice) were selected. The class of each dataset was classified by using the classification algorithms (KNN, Logistic Regression, Nave Bayes, Random Forest, and SVM) selected in the study. For each dataset, we applied 10-fold cross validation method. 10% to 100% oversampling method is applied for each fold and meta-features of the dataset is measured. The meta-features selected are HHI, Number of Classes, Number of Features, Entropy, Reverse ReLU Silhouette Score, Nonlinearity of Linear Classifier, Hub Score. F1-score was selected as the dependent variable. As a result, the results of this study showed that the six meta-features including Reverse ReLU Silhouette Score and HHI proposed in this study have a significant effect on the classification performance. (1) The meta-features HHI proposed in this study was significant in the classification performance. (2) The number of variables has a significant effect on the classification performance, unlike the number of classes, but it has a positive effect. (3) The number of classes has a negative effect on the performance of classification. (4) Entropy has a significant effect on the performance of classification. (5) The Reverse ReLU Silhouette Score also significantly affects the classification performance at a significant level of 0.01. (6) The nonlinearity of linear classifiers has a significant negative effect on classification performance. In addition, the results of the analysis by the classification algorithms were also consistent. In the regression analysis by classification algorithm, Naïve Bayes algorithm does not have a significant effect on the number of variables unlike other classification algorithms. This study has two theoretical contributions: (1) two new meta-features (HHI, Reverse ReLU Silhouette score) was proved to be significant. (2) The effects of data characteristics on the performance of classification were investigated using meta-features. The practical contribution points (1) can be utilized in the development of classification algorithm recommendation system according to the characteristics of datasets. (2) Many data scientists are often testing by adjusting the parameters of the algorithm to find the optimal algorithm for the situation because the characteristics of the data are different. In this process, excessive waste of resources occurs due to hardware, cost, time, and manpower. This study is expected to be useful for machine learning, data mining researchers, practitioners, and machine learning-based system developers. The composition of this study consists of introduction, related research, research model, experiment, conclusion and discussion.

• Journal of the Korean Geotechnical Society
• /
• v.28 no.5
• /
• pp.13-25
• /
• 2012

Ground anchor method is one of the most popular reinforcing technology for slope in Korea. For the health monitoring of slope which is reinforced by permanent anchor for a long period, monitoring of the tension force of ground anchor is very important. However, since electromechanical sensors such as strain gauge and V/W type load cell are also subject to long-term risk as well as suffering from noise during long distance transmission and immunity to electromagnetic interference (EMI), optical FBG sensors embedded tendon was developed to measure strain of 7-wire strand by embedding FBG sensor into the center king cable of 7-wire strand. This FBG sensors embedded tendon has been successfully applied to measuring the short-term anchor force. But to adopt this tendon to long-term monitoring, temperature compensation of the FBG sensors embedded tendon should be done. In this paper, we described how to compensate the effect in compliance with the change of underground temperature during long-term tension force monitoring of ground anchors by using optical fiber sensors (FBG: Fiber Bragg Grating). The model test was carried out to determine the temperature sensitivity coefficient (${\beta}^{\prime}$) of FBG sensors embedded tendon. The determined temperature sensitivity coefficient ${\beta}^{\prime}=2.0{\times}10^{-5}/^{\circ}C$ was verified by comparing the ground temperatures predicted from the proposed sensor using ${\beta}^{\prime}$ with ground temperatures measured from ground thermometer. Finally, temperature compensations were carried out based on ${\beta}^{\prime}$ value and ground temperature measurement from KMA for the tension force monitoring results of tension type and compression type anchors, which had been installed more than 1 year before at the test site. Temperature compensated tension forces are compared with those measured from conventional load cell during the same measuring time. Test results show that determined temperature sensitivity coefficient (${\beta}^{\prime}$) of FBG sensors embedded tendon is valid and proposed temperature compensation method is also appropriate from the fact that the temperature compensated tension forces are not dependent on the change of ground temperature and are consistent with the tension forces measured from the conventional load cell.

• 한국해양학회지
• /
• v.21 no.1
• /
• pp.13-24
• /
• 1986

The daily net primary production by phytoplankton in the southeastern sea of Korea in October 1985 ranged from 0.7 to 2.7 gCm$\^$-2/ d$\^$-1/ and averaged to be 1.3 gCm$\^$-2/ d$\^$-1/. Surface total chlorophyll ranged from 0.97 to 3.59mg chlm$\^$-3/. Primary production by nano-phytoplankton(〈20$\mu\textrm{m}$) ranged from 43 to 97% in the surface layer. Optimum light intensity(Iopt)was around 300 to 700${\mu}$Es$\^$-1/m$\^$-1/. Surface primary production from 9:00 to 15:00 h was evidently inhibited by strong light intensity beyond the Iopt. Phytoplankton near the base of euphotic zone(30-40m) showed extremely low Iopt suggesting adaptation to a low light environment. Since Iopt represents the history of light experience of phytoplankton at a given depth, the extent of variation in I of phytoplankton at different depth seems to be related to the in tensity of turbulence mixing in the surface mixed layer. From the present study, ammonium excretion by macrozooplankton (〉350$\mu\textrm{m}$) contributes from 3 to 19% of daily total nitrogen requirement by phytoplandton in this area. Calculation of upward flux of nitrate to the surface mixed layer from the lower layer, based on the simple diffusion model, approximates 3% of nitrogen requirement by phytoplankton. However, large portion of nitrogen requirement by phytoplankton remains unexplained in this area. In upwelling area near the coast, adjective flux might be the major source for the nitrogen requirement by phytoplankton. This study suggests that the major nitrogen source for the phytoplankton growth might come from the pelagic regeneration by nano-and micro-sized heterotrophic plandkon. Enhancement of primary production during the passage of the warm Tsushima Current is discussed in relation with nutrient dynamics and hydrlgraphic processes in this area.

• Korean Journal of Cognitive Science
• /
• v.28 no.1
• /
• pp.1-26
• /
• 2017

This study investigated the cortical generators of P300 in college students with schizotypal traits by using an auditory oddball paradigm, event-related potentials (ERPs) and standardized low resolution brain electromagnetic tomography (sLORETA) model. We also investigated the relationship between the current density of P300 and the clinical symptoms of schizophrenia. Based on the scores of Schizotypal Personality Questionnaire(SPQ), schizotypal trait (n=37) and control (n=42) groups were selected. For the measurement of P300, an auditory oddball paradigm, in which frequent standard tones (1000Hz) and rare target tones (1500Hz) were presented randomly, was used. Participants were required to count the number of the target tones during the task and report this at the end of the experiment. The two groups did not differ significantly in the accuracy of the oddball task. The schizotypal trait group showed significantly smaller P300 amplitudes than control group. In terms of source localization, both groups showed the P300 current density over bilateral frontal, parietal, temporal and occipital lobes. However, the schizotypal trait group showed significantly reduced activations in the left superior temporal gyrus and the right middle temporal gyrus, but increased activations in both left inferior frontal gyrus and right superior frontal gyrus compared to the control group. Furthermore, a negative correlation between the current density of the right superior frontal gyrus and SPQ disorganization score was found in the schizotypal trait group. These findings indicate that the individuals with schizotypal traits have dysfunctions of frontal and temporal areas, which are known to be the source of P300, as observed in patients with schizophrenia. In addition, the present results indicate that the disorganization score, rather than total score, of the SPQ is useful in predicting the risk of future schizophrenia.

• Korean Journal of Remote Sensing
• /
• v.27 no.4
• /
• pp.467-480
• /
• 2011

Spatial and temporal variabilities of NPP(Net Primary Production) retrieved from two satellite instruments, AVHRR(Advanced Very High Resolution Radiometer, 1981-2000) and MODIS(MODerate-resolution Imaging Spectroradiometer, 2000-2006), were investigated. The range of mean NPP from A VHRR and MODIS were estimated to be 894-1068 $g{\cdot}C/m^2$/yr and 610-694.90 $g{\cdot}C/m^2$/yr, respectively. The discrepancy of NPP between the two instruments is about 325 $g{\cdot}C/m^2$/yr, and MODIS product is generally closer to the ground measurement than AVHRR despite the limitation in direct comparison such as spatial resolution and vegetation classification. The higher NPP values over South Korea are related to the regions with higher biomass (e.g., mountains) and higher annual temperature. The interannual NPP trends from the two satellite products were computed, and both mean annual trends show continuous NPP increase; 2.14 $g{\cdot}C/m^2$/yr from AVHRR(1981-2000) and 6.08 $g{\cdot}C/m^2$/yr from MODIS (2000-2006) over South Korea. Specifically, the higher increasing trends over the Southwestern region are likely due to the increasing productivity of crop fields from sufficient irrigation and fertilizer use. The retrieved NPP shows a closer relationship between monthly temperature and precipitation, which results in maximum correlation during summer monsoons. The difference in the detection wavelength and model schemes during the retrieval can make a significant difference in the satellite products, and a better accuracy in the meterological and land use data and modeling applications will be necessary to improve the satellite-based NPP data.

• The Journal of Korean Academy of Prosthodontics
• /
• v.47 no.3
• /
• pp.348-355
• /
• 2009

Statement of problem: Although the use of colorimeter to improve the accuracy of tooth color measurement has increased, it is real situation that there is almost no comparative study about the accuracy among various colorimeters. Purpose: The purpose of this study is to compare the accuracy of porcelain color reproducibility using ShadeEye-Ncc dental chroma meter (Shofu Inc., Kyoto, Japan) and Shadepilot$^{TM}$ system (Degudent Inc., Hanau, Germany). Material and methods: Color of 2M2 and 3M2 shade tap of Vitapan 3D Master shade guide were measured using the above two colorimeters. Porcelain specimens were fabricated through this data. The porcelain systems used were Vintage Halo Porcelain system with ShadeEye-Ncc dental chroma meter, and VitaOmega900 Porcelain system with Shadepilot$^{TM}$ system. Colors of shade tap models and porcelain specimen were measured with spectrophotometer, and calculated delta E based on the results. Results: The following results were obtained. 1. Shadepilot$^{TM}$ system showed more suitable porcelain color reproducibility compared to ShadeEye-Ncc dental chroma meter (P <.05). 2. The result showed the difference of porcelain color reproducibility according to shade of Shade tap which is model. More suitable porcelain color reproducibility was shown at 3M2 shade tap than at 2M2 shade tap (P <.05). 3. The average of ${\Delta}E*$ of porcelain specimens which was made by the two colorimeters was 4.44-6.14, so the difference of shade was shown through the naked eye. Conclusion: Shadepilot$^{TM}$ system showed more suitable porcelain color reproducibility. When using ShadeEye-Ncc dental chromameter, it will be necessary to improve accuracy through coloring various parts of the whole tooth.

• Journal of The Korean Association For Science Education
• /
• v.33 no.7
• /
• pp.1367-1384
• /
• 2013

The purpose of this study was to develop a Science Academic Emotion Scale for Elementary Students. To make a scale, authors extract a core of 14 emotions related to science learning situations from Kim & Kim (2013) and literature review. Items on the scale consisted of 14 emotions and science learning situations. The first preliminary scale had 174 items on it. The number of 174 items was reduced and elaborated on by three science educators. Authors verified the scale using exploratory factor analysis, confirmatory factor analysis, inter-item consistency and concurrent validity. The second preliminary scale consisted of 141 items. The preliminary scale was reduced to seven factors and 56 items by applying exploratory factor analysis twice. The seven factors include: enjoyment contentment interest, boredom, shame, discontent, anger, anxiety, and laziness. The 56 items were elaborated on by five science educators. The scale with 56 items was fixed with seven factors and 35 items to get the final scale by applying confirmatory factor analysis twice. Except for Chi-square and GFI (Goodness of Fit Index), other various goodness of fit characteristics of the seven factors and 35 items model showed good estimated figures. The Cronbach of the scale was 0.85. The Cronbach of seven factors are 0.95 in enjoyment contentment interest, 0.81 in boredom, 0.87 in shame, 0.82 in discontent, 0.87 in anger, 0.77 in anxiety, 0.81 in laziness. The correlation coefficient was 0.59 in enjoyment contentment interest, 0.54 in anxiety, 0.42 in shame, and 0.28 in boredom, which were estimated using the Science Academic Emotion Scale and National Assessment System of Science-Related Affective Domain (Kim et al., 1998). Based on the results, authors judged that the Science Academic Emotion Scale for Elementary Students achieved an acceptable validity and reliability.

• Journal of the Korean Institute of Traditional Landscape Architecture
• /
• v.30 no.2
• /
• pp.64-71
• /
• 2012

This is a basic study for recovering original form of Nongsanbyeoleop(農山別業) in Gangjin, Jeonnam, created in the latter part of Joseon period and the estimation of originla form at the time of creation was conducted by analyzing related literature and inspecting the actual site. "Joseokruki(朝夕樓記)" of Dasan Jung, Yak Yong could estimate spatial structure and using form of Nongsanbyeoleop and the arrangement of spatial structure in literature could be confirm by on-the-site inspection. The results of this study are as follows. The first, Nongsanbyeoleop managed spatial factors applying natural topography. For the spatial characteristics of Nongsanbyeoleop, the location of ancestral ritual space including deceased father's tomb and tomb house far from the main levee of Yun, Kwang Taek, a father of Yun, Seo Yu by 1.9km and housekeeping could be confirmed. The second, spatial estimation by "Joseokruki" could be possible. "Joseokruki" describes Joseokru.Youngmojae.Hanokkwan.Cheokyunjung.Sangam as construction factors, Wundang.Kookdan.Nokwunoh. as plant factors, Sookyunggan.Keumkoji.Nokeumjung.Uijanghae as hydroponic factors and Pyoeunkok.Aengjakang as natural topography factors. However, most of them were disappeared and at present, only Youngmojae, Keumgoji, Kukdan and Wundang show the past trace. The third is for the changed space of Nongsanbyeoleop and its reason. The surrounding space of Nongsanbyeoleop was planated by land arrangement in 1960s and it played a role of topographical damage because it's recognized as the plane factor including Nongsanbyeoleop's surrounding landscape rather than dotted factor. The forth, the actual measurement of Nongsanbyeoleop and digitalization of manual map of numerical value are judged to be sufficient to apply as the basic material for recovering garden in the future. Because of the diatahce changing method applied at that time, the garden recovery of Nongsanbyeoleop intended to be concreted and 3D model established by digitalized basic materials is considered to apply for multilateral studying. Thus, Nongsanbyeoleop which is byeolseo including the tomb of deceased father based on the conceptual hyo thought shows clear differences from the organized factors of Byolseowonrim of precedent studies and the importance of Byolseowonrim is sufficient. But, the constructional factors which cannot know disappeared spatial factors and accurate location became the limitation of this study. In the future, clear verification of original form must be progressed by excavation which can confirm the location of construction factors.

• Journal of the Korean Society of Safety
• /
• v.10 no.3
• /
• pp.81-95
• /
• 1995

Now, safety assurance in construction sites should be accomplished by its own organization rather than control of the code or government. It is believed that the safety assurance can be considerably improved by a lecture or an education using the existing theories or literatures up to now, but it is thought that fundamental safety assurance we not able to be accomplished without developing safety devices '||'&'||' equipment or taking fundamental measures, based on the result analyzed from workers behaviors. There are various behaviors of the workers showed in construction site, but only tests for hammerusing works such as form, re-bar, stone workers directly related to the grip strength are mainly performed, investigated and measured here for the study. The above works are similar to power grip, 7th picture on seven items which are categorized for hand grip types(Ammermin 1956 ; Jones ; Kobrick 1958). Measurements of grip strength are commonly taken in anthropometric surveys. They are easy to administer but unfortunately it is rather dubious whether they yield any data that are of interest to the engineer. Very fewer controls of tools are grasped and squeesed studies showed very little overall correlation between grip strength and other measures of bodily strength (Laubach, Kromer, and Thordsen 1972), but hammer-using work which is practically progressed in construction site are mainly influenced with grip strength. According to the investigation on work measurement, it is shown that 77% of form worker are using hammer to be related to grip strength. In this study, it is particularly noticed that wearing safety gloves in construction site is required for workers safety but 20% difference between grip strength with safety gloves and without ones are commonly neglected in the site(Fig. 1). Nevertheless, safety operation with consideration of the above 20% difference is not considered in the construction site. Factors of age, kinds of work, working time, with or without safety gloves are in vestigated '||'&'||' collected at the sites for this study. Test, not at each working hour but at 14 : 00 when the almost all of the workers think the most tired, resulting from the questionaires, also when it is shown on the research report has been performed and compared for main kinds of works : form '||'&'||' re-bar work. Tests were performed with both left SE rightand of the workers simultaneously in construction site using Rand Dynamometer(Model 78010, Lafayette Instrument Co., Indiana, U.S.A) by reading grip strength on the gauge while they are pulling, and then by interviewing on their ages, works, experiences and etc., directly. The above tests have been performed for the dates of 15th march-26th May '95 with consideration of site condition. And even if various factors of ambient temperature on the testing date, working condition, individual worker's habit and worker's condition of the previous ate are concerned with the study. Those are considered as constants in this study. Samples are formwork 53, rebar 62, electrician 5, plumber 4, welding 1 from D construction Co., Ltd, ; formwork 12, re-bar 5, electrician 2, from S construction Co., Ltd, , formwork 78, re-bar 18, plumber 31, electrician 13, labor 48, plumber 31, plasterer 15, concrete placer 6, water proof worker 3, maisony 5 from B construction Co., Ltd. As In the previously mentioned, main aspect to be investigated in this study will be from '||'&'||' re-bar work because grip strength will be directly applied to these two kinds of works ; form '||'&'||' re-bar work, eventhough there are total 405 samples taken. It is thought that a frequency of accident occurrence will be mainly two work postures "looking up '||'&'||' looking down" to be mainly sorted, but this factor is not clarified in this study because It will be needed a lot of work more. Tests has been done at possible large scale of horizontally work-extended sites within one hour in order to prevent or decrease errors '||'&'||' discrepancies from time lag of the test. Additionally, the statistical package computer program SPSS PC＋has been used for the study.

• Science of Emotion and Sensibility
• /
• v.15 no.3
• /
• pp.389-398
• /
• 2012

Virtual worlds have pursued reality as if they actually exist. In order to evaluate the sense of reality in the computer-simulated worlds, several subjective questionnaires, which include specific independent variables, have been proposed in the literature. However, the questionnaires lack reliability and validity necessary for defining and measuring the virtual realization. Few studies have been conducted to investigate the effect of visual factors on the sense of reality experienced by exposing to a virtual environment. Therefore, this study was aimed at reinvestigating the variables and proposing a more reliable and advisable questionnaire for evaluating the virtual realization, focusing on visual factors. Twenty-one questions were gleaned from the literature and subjective interviews with focused groups. Exploratory factor analysis with oblique rotation was performed on the data obtained from 200 participants(females: 100) after exposing to a virtual character image described in an extreme way. After removing poorly loading items, remained subsets were subjected to confirmatory factor analysis on the data obtained from the same participants. As a result, 3 significant factors were determined to efficiently measure the virtual realization. The determined factors included visual presence(3 subset items), visual immersion(7 subset items), and visual interactivity(4 subset items). The proposed factors were verified by conducting a subjective evaluation in which participants were asked to evaluate a 3D virtual eyeball model based on the visual presence. The results implicated that the measurement method was suitable for evaluating the degree of the virtual realization. The proposed method is expected to reasonably measure the degree of the virtual realization.