• Journal of Intelligence and Information Systems
• /
• v.20 no.4
• /
• pp.121-139
• /
• 2014

Predicting corporate failure has been an important topic in accounting and finance. The costs associated with bankruptcy are high, so the accuracy of bankruptcy prediction is greatly important for financial institutions. Lots of researchers have dealt with the topic associated with bankruptcy prediction in the past three decades. The current research attempts to use ensemble models for improving the performance of bankruptcy prediction. Ensemble classification is to combine individually trained classifiers in order to gain more accurate prediction than individual models. Ensemble techniques are shown to be very useful for improving the generalization ability of the classifier. Bagging is the most commonly used methods for constructing ensemble classifiers. In bagging, the different training data subsets are randomly drawn with replacement from the original training dataset. Base classifiers are trained on the different bootstrap samples. Instance selection is to select critical instances while deleting and removing irrelevant and harmful instances from the original set. Instance selection and bagging are quite well known in data mining. However, few studies have dealt with the integration of instance selection and bagging. This study proposes an improved bagging ensemble based on instance selection using genetic algorithms (GA) for improving the performance of SVM. GA is an efficient optimization procedure based on the theory of natural selection and evolution. GA uses the idea of survival of the fittest by progressively accepting better solutions to the problems. GA searches by maintaining a population of solutions from which better solutions are created rather than making incremental changes to a single solution to the problem. The initial solution population is generated randomly and evolves into the next generation by genetic operators such as selection, crossover and mutation. The solutions coded by strings are evaluated by the fitness function. The proposed model consists of two phases: GA based Instance Selection and Instance based Bagging. In the first phase, GA is used to select optimal instance subset that is used as input data of bagging model. In this study, the chromosome is encoded as a form of binary string for the instance subset. In this phase, the population size was set to 100 while maximum number of generations was set to 150. We set the crossover rate and mutation rate to 0.7 and 0.1 respectively. We used the prediction accuracy of model as the fitness function of GA. SVM model is trained on training data set using the selected instance subset. The prediction accuracy of SVM model over test data set is used as fitness value in order to avoid overfitting. In the second phase, we used the optimal instance subset selected in the first phase as input data of bagging model. We used SVM model as base classifier for bagging ensemble. The majority voting scheme was used as a combining method in this study. This study applies the proposed model to the bankruptcy prediction problem using a real data set from Korean companies. The research data used in this study contains 1832 externally non-audited firms which filed for bankruptcy (916 cases) and non-bankruptcy (916 cases). Financial ratios categorized as stability, profitability, growth, activity and cash flow were investigated through literature review and basic statistical methods and we selected 8 financial ratios as the final input variables. We separated the whole data into three subsets as training, test and validation data set. In this study, we compared the proposed model with several comparative models including the simple individual SVM model, the simple bagging model and the instance selection based SVM model. The McNemar tests were used to examine whether the proposed model significantly outperforms the other models. The experimental results show that the proposed model outperforms the other models.

• Journal of KIISE:Computer Systems and Theory
• /
• v.29 no.6
• /
• pp.307-317
• /
• 2002

In this paper we propose a new mesh reconstruction scheme that produces a displaced subdivision surface directly from unorganized points. The displaced subdivision surface is a new mesh representation that defines a detailed mesh with a displacement map over a smooth domain surface, but original displaced subdivision surface algorithm needs an explicit polygonal mesh since it is not a mesh reconstruction algorithm but a mesh conversion (remeshing) algorithm. The main idea of our approach is that we sample surface detail from unorganized points without any topological information. For this, we predict a virtual triangular face from unorganized points for each sampling ray from a parameteric domain surface. Direct displaced subdivision surface reconstruction from unorganized points has much importance since the output of this algorithm has several important properties: It has compact mesh representation since most vertices can be represented by only a scalar value. Underlying structure of it is piecewise regular so it ran be easily transformed into a multiresolution mesh. Smoothness after mesh deformation is automatically preserved. We avoid time-consuming global energy optimization by employing the input data dependant mesh smoothing, so we can get a good quality displaced subdivision surface quickly.

• Journal of Technology Innovation
• /
• v.24 no.4
• /
• pp.277-307
• /
• 2016

Since 2011, DCF(Discounted Cash Flow) method has been used initiatively for valuating R&D technology assets in the agricultural food industry and recently technology valuation based on royalties comparison among technology transfer transactions has been also carried out in parallel when evaluating the technology assets such as new seed development technologies. Since the DCF method which has been known until now has many input variables to be estimated, sophisticated estimation has been demanded at the time of technology valuation. In addition, considering more similar trading cases when applying sales transaction comparison or industry norm method based on information of technology transfer royalty, it is an important issue that should be taken into account in the same way in the Agri-Food industry. The main input variables used for technology valuation in the Agri-Food industry are life cycle of technology asset, the financial information related to the Agri-Food industry, discount rate, and technology contribution rate. The latest infrastructure building and data updating related to technology valuation has been carried out on a regular basis in the evaluation organization of the Agri-Food segment. This study verifies the key variables that give the most important impact on the results for the existing technology valuation in the Agri-Food industry and clarifies the difference between the existing valuation result and the outcome by referring the support information that is derived through the latest input information applied in DCF method. In addition, while presenting the scheme to complement fragment information which the latest input data just influence result of technology valuation, we tried to perform comparative analysis between the existing valuation results and the evaluated outcome after the latest of reference data for making a decision the input values to be estimated in DCF. To perform these analyzes, it was first selected the representative cases evaluated past in the Agri-Food industry, applied a sensitivity analysis for input variables based on these selected cases, and then executed a simulation analysis utilizing the key input variables derived from sensitivity analysis. The results of this study is to provide the information which there are the need for modernization of the data related to the input variables that are utilized during valuating technology assets in the Agri-Food sector and for building the infrastructure of the key input variables in DCF. Therefore it is expected to provide more fruitful information about the results of valuation.

• Radiation Oncology Journal
• /
• v.12 no.2
• /
• pp.165-173
• /
• 1994

Purpose : Metastatic cancer to the brain is a major problem for the patients with bronchogenic carcinoma, and most of these patients have a limited survival expectancy. To increase tumor control and / or to decrease late morbidity with possible shortening in over-all treatment period, multiple daily fraction technique for brain metastasis was performed. The author reperesented the results of accelerated fractionation radiotherapy in patients with brain metastases from non-small cell lung cancer. Materals and Methods : Twenty-six patients with brain metastases from non-small cell lung cancer between 1991 and 1993 received brain radiotherapy with a total dose of 48 Gy, at 2 Gy per fraction, twice a day with a interfractional period of 6 hours, and delivered 5 days a week. The whole brain was treated to 40 Gy and boost dose escalated to 8 Gy for single metastatic lesion by reduced field. Twenty-four of the 26 patients completed the radiotherapy. Radiotherapy was interupted in two patients suggesting progressive intracerebral diseases. Results : This radiotherapy regimen appears to be comparable to the conventional scheme in relief from symptoms. Three of the 24 patients experienced nausea and or vomiting during the course of treatment because of acute irradiation toxicity. The author observed no excessive toxicity with escalating dose of irradiation. An increment in median survival, although not statistically significant(p>0.05), was noted with escalating doses(48 Gy) of accelerated fractionation(7 months) compared to conventional treatment(4.5 months). Median survival also increased in patients with brain solitary metastasis(9 months) compared to multiple extrathoracic sites(4 months), and in patients with good performance status(9 months versus 3.5 months), they were statistically significant (p<0.01). Conclusion : The increment in survival in patients with good prognostic factors such as controlled primary lesion, metastasis in brain only and good perfomance status appeared encouraging. Based on these results, a multi-institutional prospective randomized trial should be initiated to compare the twice-a-day and once-a-day radiotherapy schemes on patients with brain metastasis with careful consideration for the patients' quality of life.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.17 no.4
• /
• pp.384-398
• /
• 2015

In this paper, the high-resolution Weather Research and Forecasting/Noah-MultiParameterization (WRF/Noah-MP) modeling system is configured for the Cheongmicheon Farmland site in Korea (CFK), and its performance in land and atmospheric simulation is evaluated using the observed data at CFK during the 2014 special observation period (21 August-10 September). In order to explore the usefulness of turning on Noah-MP dynamic vegetation in midterm simulations of surface and atmospheric variables, two numerical experiments are conducted without dynamic vegetation and with dynamic vegetation (referred to as CTL and DVG experiments, respectively). The main results are as following. 1) CTL showed a tendency of overestimating daytime net shortwave radiation, thereby surface heat fluxes and Bowen ratio. The CTL experiment showed reasonable magnitudes and timing of air temperature at 2 m and 10 m; especially the small error in simulating minimum air temperature showed high potential for predicting frost and leaf wetness duration. The CTL experiment overestimated 10-m wind and precipitation, but the beginning and ending time of precipitation were well captured. 2) When the dynamic vegetation was turned on, the WRF/Noah-MP system showed more realistic values of leaf area index (LAI), net shortwave radiation, surface heat fluxes, Bowen ratio, air temperature, wind and precipitation. The DVG experiment, where LAI is a prognostic variable, produced larger LAI than CTL, and the larger LAI showed better agreement with the observed. The simulated Bowen ratio got closer to the observed ratio, indicating reasonable surface energy partition. The DVG experiment showed patterns similar to CTL, with differences for maximum air temperature. Both experiments showed faster rising of 10-m air temperature during the morning growth hours, presumably due to the rapid growth of daytime mixed layers in the Yonsei University (YSU) boundary layer scheme. The DVG experiment decreased errors in simulating 10-m wind and precipitation. 3) As horizontal resolution increases, the models did not show practical improvement in simulation performance for surface fluxes, air temperature, wind and precipitation, and required three-dimensional observation for more agricultural land spots as well as consistency in model topography and land cover data.

• Journal of Elementary Mathematics Education in Korea
• /
• v.11 no.2
• /
• pp.177-197
• /
• 2007

This study discussed the climbing learning method which studied and practiced by Professor Saito Noboru. This is the learning method which is devised to know not only the relationship of the learning factors but the systemic or structural connection of whole studying contents- affects children's math learning ability through practical class to both the lower and the higher grades. To achieve the purpose of this study, these following issues were set; A. Develop the teaching and learning course of mathematics by applying the climbing learning method. B. Execute the mathematics lesson according to the climbing learning method and analyze the learning achievement. C. Analyze the difference between application of the climbing learning method and that of the learning method by student's level in mathematics. D. Analyze what the climbing learning method gives a shift of the recognition of learning mathematics. In order to accomplish these study issues, we analyzed the text book of math not only for children but also for teachers and developed the teaching and learning course applied the climbing learning method with advice of experts. It was chosen two different homogeneous groups each, third year for lower grade group and fifth year for higher grade group. It was done the experimental group lesson applying the climbing learning method and general lesson for the control group. After then, t-test against independent samples was done depending on the result of the student's assessment(T1, T2). These two groups' students were divided into smaller groups based on result of achievement level regardless of gender. These subgroups were confirmed the difference of learning ability between upper and lower level group. As regarding the result making out grades of faith and attitude for math, t-test was used on independent sample. At the same time, experimental groups were tested using learning attitude with the learning structure chart. Through this study the following results are obtained and the conclusion was drawn. Firstly, although applying the climbing learning method to the lesson does not have significant effect to the lower grade of elementary school student's achievement it has significant influence on the higher grade student's achievement. Second, as a result of analyzing the difference between the climbing learning method and the learning method by student's level in mathematics, it is of no beneficial effect to the lower grade both upper level and lower level. However, it has appreciable effect to the higher grade classes both upper level and low level. Especially, upper level students have higher effect than low level students. Third, climbing learning method does not affect to the faith and attitude of the lower grade students positively, but it has affirmative effect to the higher grade students'. As a result of the survey of the experimental groups which were applied to the climbing loaming method, the lesson by using the learning structure chart proved to be helpful to the both the lower and higher grade. The best advantage of using the learning structure chart, children say, is easily understood whole contents of studying and is useful for review. Furthermore, using the learning structure chart is more efficient compared with previous learning method and is given the successful result to self-directed learning. In conclusion, keeping up with the current of the thought of education, we suggest a scheme as a new teaching method from the constructive learning method which emphasize the self-directed learning.

• Journal of the Korean association of regional geographers
• /
• v.14 no.3
• /
• pp.269-278
• /
• 2008

This research analyzed the effect of lessons with the GIS application as an alternative scheme of teaching and learning of geography in elementary school. Two classes in the third grade at Y elementary school in Andong were selected to conduct lessons on 'The Landscape of My Hometown' from March 6 through June 30, 2006. In the experimental class, the lessons were conducted with the GIS application; while, in a comparative class, the lessons were carried with usual teaching and learning method. To find out the effect of lessons with the GIS application, differences of spatial cognition of students were figured out between groups, and before and after lessons. The difference between the spatial concept development stages and materials on the textbook discouraged students to pursue their learning as well as made them hard to achieve the goals of lessons. The GIS application had been suggested as an alternative teaching and learning method to overcome the difference; however, it has been hard to find any empirical research to verify the effect of the lessons with GIS application in elementary school. The ability of spatial cognition of the third graders at an elementary school was very low as the result of that curricula in the first and second grades dealt with sketch maps as teaching and learning media. The map learning of third grader on the transitional stage would play the critical role to develop the spatial cognition ability in the future. The field study contributing to developing spatial cognition ability would not be conducted at school. It was required to have the alternative learning schemes such as lessons with GIS application. The lessons with GIS application verified effect of GIS application as the alternative method. The GIS application helped students to recognize landmarks, directions and distance effectively as well as reduced the spatial cognition difference among individuals and/or groups.

• Journal of KIISE:Information Networking
• /
• v.35 no.2
• /
• pp.99-111
• /
• 2008

In recent years, the needs for WLANs(Wireless Local Area Networks) technology which can access to Internet anywhere have been dramatically increased particularly in SOHO(Small Office Home Office) and Hot Spot. However, unlike wired networks, there are some unique characteristics of wireless networks. These characteristics include the burst packet losses due to unreliable wireless channel. Note that burst packet losses, which occur when the distance between the wireless station and the AP(Access Point) increase or when obstacles move temporarily between the station and AP, are very frequent in 802.11 networks. Conversely, due to burst packet losses, the performance of 802.11 networks are not always as sufficient as the current application require, particularly when they use TCP at the transport layer. The high packet loss rate over wireless links can trigger unnecessary execution of TCP congestion control algorithm, resulting in performance degradation. In order to overcome the limitations of WLANs environment, MAC-layer LDA(Loss Differentiation Algorithm)has been proposed. MAC-layer LDA prevents TCP's timeout by increasing CRD(Consecutive Retry Duration) higher than burst packet loss duration. However, in the wireless channel with high packet loss rate, MAC-layer LDA does not work well because of two reason: (a) If the CRD is lower than burst packet loss duration due to the limited increase of retry limit, end-to-end performance is degraded. (b) energy of mobile device and bandwidth utilization in the wireless link are wasted unnecessarily by Reducing the drainage speed of the network buffer due to the increase of CRD. In this paper, we propose a new retransmission module based on Cross-layer approach, called BLD(Burst Loss Detection) module, to solve the limitation of previous link layer retransmission schemes. BLD module's algorithm is retransmission mechanism at IEEE 802.11 networks and performs retransmission based on the interaction between retransmission mechanisms of the MAC layer and TCP. From the simulation by using ns-2(Network Simulator), we could see more improved TCP throughput and energy efficiency with the proposed scheme than previous mechanisms.

• Journal of Korean Society of Environmental Engineers
• /
• v.38 no.11
• /
• pp.603-610
• /
• 2016

This work studies the synthesis of birnessite (${\delta}-MnO_2$), a catalyst of oxidative-coupling reactions, from the powder of spent alkaline manganese batteries (SABP, <8 mesh) and evaluate its reactivity for 1-naphthol (1-NP) removals. Manganese oxides using commercial reagents ($MnSO_4$, $MnCl_2$) and the acid birnessite (A-Bir) by McKenzie method were also synthesized, and their crystallinity and reactivity for 1-NP were compared with one another. 96% Mn and 98% Zn were extracted from SABP by acid leaching at the condition of solid/liquid (S/L) ratio 1:10 in $1.0M\;H_2SO_4+10.5%\;H_2O_2$ at $60^{\circ}C$. From the acid leaching solution, 69% (at pH 8) and 94.3% (pH>13) of Mn were separated by hydroxide precipitation. Optimal OH/Mn mixing ratio (mol/mol) for the manganese oxide (MO) synthesis by alkaline (NaOH) hydrothermal techniques was 6.0. Under this condition, the best 1-NP removal efficiency was observed and XRD analysis confirmed that the MOs are corresponding to birnessite. Kinetic constants (k, at pH 6) for the 1-NP removals of the birnessites obtained from Mn recovered at pH 8 (${Mn^{2+}}_{(aq)}$) and pH>13 ($Mn(OH)_{2(s)}$) are 0.112 and $0.106min^{-1}$, respectively, which are similar to that from $MnSO_4$ reagent ($0.117min^{-1}$). The results indicated that the birnessite prepared from the SABP as a raw material could be used as an oxidative-coupling catalyst for removals of trace phenolic compounds in soil and water, and propose the recycle scheme of SAB for the birnessite synthesis.

• Journal of Astronomy and Space Sciences
• /
• v.26 no.4
• /
• pp.425-438
• /
• 2009

We reevaluate the Burton equation (Burton et al. 1975) of predicting Dst index using high quality hourly solar wind data supplied by the ACE satellite for the period from 1998 to 2006. Sixty magnetic storms with monotonously decreasing main phase are selected. In order to determine the injection term (Q) and the decay time ($\tau$) of the equation, we examine the relationships between $Dst^*$ and $VS_s$, ${\Delta}Dst^*$ and $VS_s$, and ${\Delta}Dst^*$ and $Dst^*$ during the magnetic storms. For this analysis, we take into account one hour of the propagation time from the ACE satellite to the magnetopause, and a half hour of the response time of the magnetosphere/ring current to he solar wind forcing. The injection term is found to be $Q(nT/h)\;=\;-3.56VS_s$ for $VS_s$ > 0.5mV/m and Q(nT=h) = 0 for $VB_s\;{\leq}\;0.5mV/m$. The $\tau$ (hour) is estimated as $0.060Dst^*\;+\;16.65$ for $Dst^*$ > -175nT and 6.15 hours for $Dst^*\;{\leq}\;-175nT$. Based on these empirical relationships, we predict the 60 magnetic storms and find that the correlation coefficient between the observed and predicted $Dst^*$ is 0.88. To evaluate the performance of our prediction scheme, the 60 magnetic storms are predicted again using the models by Burton et al. (1975) and O'Brien & McPherron (2000a). The correlation coefficients thus obtained are 0.85, the same value for both of the two models. In this respect, our model is slightly improved over the other two models as far as the correlation coefficients is concerned. Particularly our model does a better job than the other two models in predicting intense magnetic storms ($Dst^*\;{< \atop \sim}\;-200nT$).