• 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.

• Radiation Oncology Journal
• /
• v.21 no.4
• /
• pp.322-329
• /
• 2003

Purpose: It is difficult to exactly determine the surface dose and the dose distribution In buildup region of high energy X-rays by using the conventional ion chamber. The aim of this study Is to evaluate the accuracy of widely used dosimetry systems to measure the surface dose and the depth of maximum dose (d$_{max}$). Materials and Methods: We measured the percent depth dose (PDD) from the surface to the d$_{max}$ in either a water phantom or in a solid water phantom using TLD-100 chips, thimble type ion chamber, diode detector, diamond detector and Markus parallel plate ion chamber for 6 MV and 15 MV X-rays, 10$\times$10 cm$^{2}$, at SSD=100cm. We analysed the surface dose and the d$_{max}$. In order to verify the accuracy of the TLD data, we executed the Monte Carlo simulation for 5 MV X-ray beams. Results: The surface doses In 6 MV and IS MV X-rays were 29.31% and 23.36% ior Markus parallel plate ion chamber, 37.17$\%$ and 24.01$\%$ for TLD, 34.87$\%$ and 24.06$\%$ for diamond detector, 38.13$\%$ and 27.8$\%$ for diode detector, and 47.92$\%$ and 35.01$\%$ for thimble type ion chamber, respectively. in Monte Carlo simulation for 6 MV X-rays, the surface dose was 36.22$\%$, which Is similar to the 37.17$\%$ of the TLD measurement data. The d$_{max}$ In 6 WV and 15 MV X-rays was 14$\~$16 mm and 27$\~$29 mm, respectively. There was no significant difference in the d$_{max}$ among the detectors. Conclusion: There was a remarkable difference in the surface dose among the detectors. The Markus parallel plate chamber showed the most accurate result. The surface dose of the thimble ion chamber was 10$\%$ higher than that of other detectors. We suggest that the correction should be made when the surface dose of the thimble ion chamber Is used for the treatment planning ion the supeficial tumors. All the detectors used In our study showed no difference in the d$_{max}$.

• Proceedings of the Korea Association of Crystal Growth Conference
• /
• 1996.06a
• /
• pp.179-200
• /
• 1996

The intrinsic instabilities of fluid flow occurred in the melt of the Czochralski crystal growth system Czochralski method, asymmetric flow patterns and temperature profiles in the melt have been studied by many researchers. The idea that the non-symmetric structure of the growing equipment is responsible for the asymmetric profiles is usually accepted at the first time. However further researches revealed that some intrinsic instabilities not related to the non-symmetric equipment structure in the melt could also appear. Ristorcelli had pointed out that there are many possible causes of instabilities in the melt. The instabilities appears because of the coupling effects of fluid flow and temperature profiles in the melt. Among the instabilities, the B nard type instabilities with no or low crucible rotation rates are analyzed by the visualizing experiments using X-ray radiography and the 3-D numerical simulation in this study. The velocity profiles in the Silicon melt at different crucible rotation rates were measured using X-ray radiography method using tungsten tracers in the melt. The results showed that there exits two types of fluid flow mode. One is axisymmetric flow, the other is asymmetric flow. In the axisymmetric flow, the trajectory of the tracers show torus pattern. However, more exact measurement of the axisymmetrc case shows that this flow field has small non-axisymmetric components of the velocity. When fluid flow is asymmetric, the tracers show random motion from the fixed view point. On the other hand, when the observer rotates to the same velocity of the crucible, the trajectory of the tracer show a rotating motion, the center of the motion is not same the center of the melt. The temperature of a point in the melt were measured using thermocouples with different rotating rates. Measured temperatures oscillated. Such kind of oscillations are also measured by the other researchers. The behavior of temperature oscillations were quite different between at low rotations and at high rotations. Above experimental results means that the fluid flow and temperature profiles in the melt is not symmetric, and then the mode of the asymmetric is changed when rotation rates are changed. To compare with these experimental results, the fluid flow and temperature profiles at no rotation and 8 rpm of crucible rotation rates on the same size of crucible is calculated using a 3-dimensional numerical simulation. A finite different method is adopted for this simulation. 50×30×30 grids are used. The numerical simulation also showed that the velocity and flow profiles are changed when rotation rates change. Futhermore, the flow patterns and temperature profiles of both cases are not axisymmetric even though axisymmetric boundary conditions are used. Several cells appear at no rotation. The cells are formed by the unstable vertical temperature profiles (upper region is colder than lower part) beneath the free surface of the melt. When the temperature profile is combined with density difference (Rayleigh-B nard instability) or surface tension difference (Marangoni-B nard instability) on temperature, cell structures are naturally formed. Both sources of instabilities are coupled to the cell structures in the melt of the Czochralski process. With high rotation rates, the shape of the fluid field is changed to another type of asymmetric profile. Because of the velocity profile, isothermal lines on the plane vertical to the centerline change to elliptic. When the velocity profiles are plotted at the rotating view point, two vortices appear at the both sides of centerline. These vortices seem to be the main reason of the tracer behavior shown in the asymmetric velocity experiment. This profile is quite similar to the profiles created by the baroclinic instability on the rotating annulus. The temperature profiles obtained from the numerical calculations and Fourier transforms of it are quite similar to the results of the experiment. bove esults intend that at least two types of intrinsic instabilities can occur in the melt of Czochralski growing systems. Because the instabilities cause temperature fluctuations in the melt and near the crystal-melt interface, some defects may be generated by them. When the crucible size becomes large, the intensity of the instabilities should increase. Therefore, to produce large single crystals with good quality, the behavior of the intrinsic instabilities in the melt as well as the effects of the instabilities on the defects in the ingot should be studied. As one of the cause of the defects in the large diameter Silicon single crystal grown by the

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.7 no.2
• /
• pp.113-127
• /
• 2012

From the end of 1st quarter of 2012, Korean mandatory firms had started releasing financial reports conforming to the K-IFRS(Korean adopted International Financial Reporting Standards). Major characteristics of IFRS, such as 'principles based' features, consolidated reporting, 'fair value' measurement, increased pressure for non-financial disclosures have resulted in brief and various disclosure practices regarding the main body of each statements and vast amount of note description requirements. Meanwhile, a host of previous studies on IFRS disclosures have incorporated regulatory and/or 'compete information' perspectives, mainly focusing on suggesting further enforcement of strengthened requirements and providing guidelines for specific treatments. Thus, as an extension of prior findings and suggestions this study had explored to conduct an integrative approach embracing views of the reporting entities and the users of financial information. In spite of all the state-driven efforts for faithful representation and comparability of corporate financial reports, an overhaul of disclosure practices of fiscal year 2010 and 2011 had revealed numerous cases of insufficiency and discordance in terms of mandatory norms and market expectations. As to the causes of such shortcomings, this study identified several factors from the corporate side and the users of the information; some inherent aspects of IFRS, industry/corporate-specific context, expenditures related to internalizing IFRS system, reduced time frame for presentation. lack of clarity and details to meet the quality of information - understandability, comparability etc. - commonly requested by the user group. In order to improve current disclosure practices, dual approach had been suggested; Firstly, to encourage and facilitate implementation, (1) further segmentation and differentiation of mandates among companies, (2) redefining the scope and depth of note descriptions, (3) diversification and coordination of reporting periods, (4) providing support for equipping disclosure systems and granting incentives for best practices had been discussed. Secondly, as for the hard measures, (5) regularizing active involvement of corporate and user group delegations in the establishment and amendment process of K-IFRS (6) enforcing detailed and standardized disclosure on reporting entities had been recommended.

• Journal of Bio-Environment Control
• /
• v.26 no.3
• /
• pp.181-187
• /
• 2017

The purpose of this study was to provide basic data for development of environmental design technology for greenhouses constructed in reclaimed lands. The climatic conditions around seven major reclaimed land areas with a plan to install advanced horticultural complexes in Korea were analyzed. The characteristics of natural ventilation and temperature rise through the thermal environment measurement of the greenhouse in Saemangeum were analyzed. The part to be applied to the environmental design of the greenhouses in reclaimed lands were reviewed. Results of comparing the ventilation rate of the greenhouse according to the presence or absence of plants showed the greenhouse with plants had the lower ventilation rate, but the smaller rise of indoor temperature due to the evapotranspiration of plants. In the greenhouse with plants, the number of air changes was in the range of 0.3 to 0.9 volumes/min and the average was 0.7 volumes/min. The rise of indoor temperature relative to outdoor temperature was in the range of 1 to $5^{\circ}C$ and the average $2.5^{\circ}C$. The natural ventilation performance of the experimental greenhouse constructed in the reclaimed land almost satisfied the recommended ventilation rate in summer and the rise of indoor temperature relative to outdoor temperature did not deviate considerably from the cultivation environment of plants. Therefore, it was determined that the greenhouse cultivation in Saemangeum reclaimed land is possible with only natural ventilation systems without cooling facilities. As the reclaimed land is located in the seaside, the wind is stronger than the inland area, and the fog is frequent. This strong wind speed increases the ventilation rate of greenhouses, which is considered to be a factor for reducing the cooling load. In addition, since the fog duration is remarkably longer than that of inland area, the seasonal cooling load is expected to decrease, which is considered to be advantageous in terms of the operation cost of cooling facilities.

• Korean Journal of Cognitive Science
• /
• v.20 no.2
• /
• pp.125-154
• /
• 2009

This paper describes the possibility of human physiological data, especially brain-wave activity, to detect cognitive overload, a phenomenon that may occur while learner uses an e-learning system. If it is found that cognitive overload to be detectable, providing appropriate feedback to learners may be possible. To illustrate the possibility, while engaging in cognitive activities, cognitive load levels were measured by EEG (electroencephalogram) to seek detection of cognitive overload. The task given to learner was a computerized listening and recall test designed to measure working memory capacity, and the test had four progressively increasing degrees of difficulty. Eight male, right-handed, university students were asked to answer 4 sets of tests and each test took from 61 seconds to 198 seconds. A correction ratio was then calculated and EEG results analyzed. The correction ratio of listening and recall tests were 84.5%, 90.6%, 62.5% and 56.3% respectively, and the degree of difficulty had statistical significance. The data highlighted learner cognitive overload on test level of 3 and 4, the higher level tests. Second, the SEF-95% value was greater on test3 and 4 than on tests 1 and 2 indicating that tests 3 and 4 imposed greater cognitive load on participants. Third, the relative power of EEG gamma wave rapidly increased on the 3rd and $4^{th}$ test, and signals from channel F3, F4, C4, F7, and F8 showed statistically significance. These five channels are surrounding the brain's Broca area, and from a brain mapping analysis it was found that F8, right-half of the brain area, was activated relative to the degree of difficulty. Lastly, cross relation analysis showed greater increasing in synchronization at test3 and $4^{th}$ at test1 and 2. From these findings, it is possible to measure brain cognitive load level and cognitive over load via brain activity, which may provide atimely feedback scheme for e-learning systems.

• Journal of the Korean Wood Science and Technology
• /
• v.37 no.1
• /
• pp.94-104
• /
• 2009

It is urgently required to develop the production of fermentation-heat energy from the waste agricultural and forest biomass and its effective heat exchanging system for the supply of warm water to rural households and greenhouses. In this study 3 helical-type and 1 plate-type heat exchangers using 3 different waste biomasses [e.g. hardwood (HW) sawdust (100%), softwood (SW) sawdust : HW sawdust (50 : 50) and HW sawdust : grass (90 : 10)] were applied in order to find out the best heat recovery system. The heat exchanger was basically considered to improve the overall heat recovery efficiency, to minimize heat loss and to simplify manufacturing, assembling and breaking up the fermenting beds. The helical-type heat exchanger (HX-H3) installed in fermenting bed of HW sawdust : grass (90 : 10) showed relatively higher temperature profiles, in particular mid- and upper-parts than lower and surface parts during 45-day fermentation process. The maximum temperature was ranged from $40^{\circ}C$ to $65^{\circ}C$ with average $60^{\circ}C$. The water temperature of tank outlet was ranged to $33{\sim}48^{\circ}C$ during whole measuring periods. By the way plate-type one (HX-P) installed in same biomass compositional fermenting bed showed $64.5{\sim}76.5^{\circ}C$ at center part, and $43{\sim}56^{\circ}C$ and $42{\sim}58^{\circ}C$, water tank and tank outlet temperatures, respectively, during 100 day measurement. It could be concluded that the plate-type heat exchanger (HX-P) provides not only the effective heating for the rural households and greenhouses, but also having the best heat recovery performance, easy manufacturing, assembling and breaking up the systems.

• Economic and Environmental Geology
• /
• v.43 no.1
• /
• pp.13-20
• /
• 2010

Acid mine drainage (AMD) from abandoned mine sites typically has low pH and contains high level of various heavy metals, aggravating ground- and surface water qualities and neighboring environments. This study investigated removal of heavy metals in a biological treatment system, mainly focusing on the removal by adsorption on a substrate material. Bench-scale batch experiments were performed with a mushroom compost to evaluate the adsorption characteristics of heavy metals leached out from a mine tailing sample and the role of SRB in the overall removal process. In addition, adsorption experiments were perform using an artificial AMD sample containing $Cd^{2+}$, $Cu^{2+}$, $Pb^{2+}$ and $Zn^{2+}$ to assess adsorption capacity of the mushroom compost. The results indicated Mn leached out from mine tailing was not subject to microbial stabilization or adsorption onto mushroom compost while microbially mediated stabilization played an important role in the removal of Zn. Fe leaching significantly increased in the presence of microbes as compared to autoclaved samples, and this was attributed to dissolution of Fe minerals in the mine tailing in a response to the depletion of $Fe^{3+}$ by iron reduction bacteria. Measurement of oxidation reduction potential (ORP) and pH indicated the reactive mixture maintained reducing condition and moderate pH during the reaction. The results of the adsorption experiments involving artificial AMD sample indicated adsorption removal efficiency was greater than 90% at pH 6 condition, but it decreased at pH 3 condition.

• 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 Intelligence and Information Systems
• /
• v.16 no.4
• /
• pp.43-65
• /
• 2010

Facing a complex environment driven by a decade, many companies are adopting new strategic frameworks such as Customer Relationship Management system to achieve sustainable profitability as well as overcome serious competition for survival. In many business areas, CRM system advanced a great deal in a matter of continuous compensating the defect and overall integration. However, pharmaceutical companies in Korea were slow to accept them for usesince they still have a tendency of holding fast to traditional way of sales and marketing based on individual networks of sales representatives. In the circumstance, this article tried to empirically address current status of CRM system as well as the effects of the system on the performance of pharmaceutical companies by applying BSC method's four perspectives, from financial, customer, learning and growth and internal process. Survey by e-mail and post to employers and employees who were working in pharma firms were undergone for the purpose. Total 113 cases among collected 140 ones were used for the statistical analysis by SPSS ver. 15 package. Reliability, Factor analysis, regression were done. This study revealed that CRM system had a significant effect on improving financial and non-financial performance of pharmaceutical companies as expected. Proposed regression model fits well and among them, CRM marketing information system shed the light on substantial impact on companies' outcome given profitability, growth and investment. Useful analytical information by CRM marketing information system appears to enable pharmaceutical firms to set up effective marketing and sales strategies, these result in favorable financial performance by enhancing values for stakeholderseventually, not to mention short-term profit and/or mid-term potential to growth. CRM system depicted its influence on not only financial performance, but also non-financial fruit of pharmaceutical companies. Further analysis for each component showed that CRM marketing information system were able to demonstrate statistically significant effect on the performance like the result of financial outcome. CRM system is believed to provide the companies with efficient way of customers managing by valuable standardized business process prompt coping with specific customers' needs. It consequently induces customer satisfaction and retentionto improve performance for long period. That is, there is a virtuous circle for creating value as the cornerstone for sustainable growth. However, the research failed to put forward to evidence to support hypothesis regarding favorable influence of CRM sales representative's records assessment system and CRM customer analysis system on the management performance. The analysis is regarded to reflect the lack of understanding of sales people and respondents between actual work duties and far-sighted goal in strategic analysis framework. Ordinary salesmen seem to dedicate short-term goal for the purpose of meeting sales target, receiving incentive bonus in a manner-of-fact style, as such, they tend to avail themselves of personal network and sales and promotional expense rather than CRM system. The study finding proposed a link between CRM information system and performance. It empirically indicated that pharmaceutical companies had been implementing CRM system as an effective strategic business framework in order for more balanced achievements based on the grounded understanding of both CRM system and integrated performance. It suggests a positive impact of supportive CRM system on firm performance, especially for pharmaceutical industry through the initial empirical evidence. Also, it brings out unmet needs for more practical system design, improvement of employees' awareness, increase of system utilization in the field. On the basis of the insight from this exploratory study, confirmatory research by more appropriate measurement tool and increased sample size should be further examined.