• Asia pacific journal of information systems
• /
• v.19 no.2
• /
• pp.139-155
• /
• 2009

For the last few decades, many studies have tried to explore and unveil venture companies' success factors and unique features in order to identify the sources of such companies' competitive advantages over their rivals. Such venture companies have shown tendency to give high returns for investors generally making the best use of information technology. For this reason, many venture companies are keen on attracting avid investors' attention. Investors generally make their investment decisions by carefully examining the evaluation criteria of the alternatives. To them, credit rating information provided by international rating agencies, such as Standard and Poor's, Moody's and Fitch is crucial source as to such pivotal concerns as companies stability, growth, and risk status. But these types of information are generated only for the companies issuing corporate bonds, not venture companies. Therefore, this study proposes a method for evaluating venture businesses by presenting our recent empirical results using financial data of Korean venture companies listed on KOSDAQ in Korea exchange. In addition, this paper used multi-class SVM for the prediction of DEA-based efficiency rating for venture businesses, which was derived from our proposed method. Our approach sheds light on ways to locate efficient companies generating high level of profits. Above all, in determining effective ways to evaluate a venture firm's efficiency, it is important to understand the major contributing factors of such efficiency. Therefore, this paper is constructed on the basis of following two ideas to classify which companies are more efficient venture companies: i) making DEA based multi-class rating for sample companies and ii) developing multi-class SVM-based efficiency prediction model for classifying all companies. First, the Data Envelopment Analysis(DEA) is a non-parametric multiple input-output efficiency technique that measures the relative efficiency of decision making units(DMUs) using a linear programming based model. It is non-parametric because it requires no assumption on the shape or parameters of the underlying production function. DEA has been already widely applied for evaluating the relative efficiency of DMUs. Recently, a number of DEA based studies have evaluated the efficiency of various types of companies, such as internet companies and venture companies. It has been also applied to corporate credit ratings. In this study we utilized DEA for sorting venture companies by efficiency based ratings. The Support Vector Machine(SVM), on the other hand, is a popular technique for solving data classification problems. In this paper, we employed SVM to classify the efficiency ratings in IT venture companies according to the results of DEA. The SVM method was first developed by Vapnik (1995). As one of many machine learning techniques, SVM is based on a statistical theory. Thus far, the method has shown good performances especially in generalizing capacity in classification tasks, resulting in numerous applications in many areas of business, SVM is basically the algorithm that finds the maximum margin hyperplane, which is the maximum separation between classes. According to this method, support vectors are the closest to the maximum margin hyperplane. If it is impossible to classify, we can use the kernel function. In the case of nonlinear class boundaries, we can transform the inputs into a high-dimensional feature space, This is the original input space and is mapped into a high-dimensional dot-product space. Many studies applied SVM to the prediction of bankruptcy, the forecast a financial time series, and the problem of estimating credit rating, In this study we employed SVM for developing data mining-based efficiency prediction model. We used the Gaussian radial function as a kernel function of SVM. In multi-class SVM, we adopted one-against-one approach between binary classification method and two all-together methods, proposed by Weston and Watkins(1999) and Crammer and Singer(2000), respectively. In this research, we used corporate information of 154 companies listed on KOSDAQ market in Korea exchange. We obtained companies' financial information of 2005 from the KIS(Korea Information Service, Inc.). Using this data, we made multi-class rating with DEA efficiency and built multi-class prediction model based data mining. Among three manners of multi-classification, the hit ratio of the Weston and Watkins method is the best in the test data set. In multi classification problems as efficiency ratings of venture business, it is very useful for investors to know the class with errors, one class difference, when it is difficult to find out the accurate class in the actual market. So we presented accuracy results within 1-class errors, and the Weston and Watkins method showed 85.7% accuracy in our test samples. We conclude that the DEA based multi-class approach in venture business generates more information than the binary classification problem, notwithstanding its efficiency level. We believe this model can help investors in decision making as it provides a reliably tool to evaluate venture companies in the financial domain. For the future research, we perceive the need to enhance such areas as the variable selection process, the parameter selection of kernel function, the generalization, and the sample size of multi-class.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2011.05a
• /
• pp.46-46
• /
• 2011

Severe sediment erosion during floods occur disaster and economic losses, but general sediment erosion is basic mechanism to move sediment from upstream to downstream river. In addition, it is important process to change river form. Check dam, which is constructed in mountain stream, play a vital role such as control of sudden debris flow, but it has negative aspects to river ecosystem. Now a day, check dam of open type is an alternative plan to recover river biological diversity and ecosystem through sediment transport while maintaining the function of disaster control. The purpose of this paper is to verify sediment erosion progress of river bottom and bank as first step for river restoration after dam slit by cross-sectional shear stress and critical shear stress. Study area is upstream reach of slit check dam in mountain stream, named Wasada, in Japan. The check dam was slit with two passages in August, 2010. The transects were surveyed for four upstream cross-sections, 7.4 m, 34 m, 86 m, and 150 m distance from dam in October 2010. Sediment size was surveyed at river bottom and bank. Sediment of cobble size was found at the wetted bottom, and small size particles of sand to medium gravel composed river bank. Discharge was $2.5\;m^3/s$ and bottom slope was 0.027 m/m. Excess shear stress (${\tau}_{ex}$) was calculated for hydraulic erosion by subtracting the values of critical shear stress (${\tau}_{c}$) from the value of shear stress (${\tau}$) at river bottom and bank (${\tau}_{ex}=\tau-{\tau}_c$). Shear stress of river bottom (${\tau}_{bottom}$) was calculated using the cross-sectional shear stress, and bank shear stress (${\tau}_{bank}$) was calculated from the method of Flintham and Carling (1988). $${\tau}_{bank}={\tau}^*SF_{bank}((B+P_{bed})/(2^*P_{bank}))$$ where $SF_{bank}=1.77(P_{bed}/p_{bank}+1.5)^{-1.4}$, B is the water surface width, $P_{bed}$ and $P_{bank}$ are wetted parameter of the bed and bank. Estimated values for ${\tau}_{bottom}$ for a flow of $2.5\;m^3/s$ were lower as 25.0 (7.5 m cross-section), 25.7 (34 m), 21.3 (86 m) and 19.8 (150 m), in N/$m^2$, than critical shear stress (${\tau}_c=62.1\;N/m^2$) with cobble of 64 mm. The values were insufficient to erode cobble sediment. In contrast, even if the values of ${\tau}_{bank}$ were lower than the values for ${\tau}_{bottom}$ as 18.7 (7.5 m), 19.3 (34 m), 16.1 (86 m) and 14.7 (150 m), in N/$m^2$, excess shear stresses were calculated at the three cross-sections of 7.5 m, 34 m, and 86 m distances compare with ${\tau}_c$ is 15.5 N/$m^2$ of 16mm gravel. Bank shear stresses were sufficient for erosion of the medium gravel to sand. Therefore there is potential to erode lateral bank than downward erosion in a flow of $2.5\;m^3/s$. Undercutting of the wetted bank can causes bank scour or collapse, therefore this channel has potential to become wider at the same time. This research is about a potential of sediment erosion, and the result could not verify with real data. Therefore it need next step for verification. In addition an erosion mechanism for river restoration is not simple because discharge distribution is variable by snow-melting or rainy season, and a function for disaster control will recover by big precipitation event. Therefore it needs to consider the relationship between continuous discharge change and sediment erosion.

• Tuberculosis and Respiratory Diseases
• /
• v.45 no.1
• /
• pp.116-127
• /
• 1998

Background: The natural course of sarcoidosis is variable from spontaneous remission to significant morbidity or death. So the assessment of disease activity is important but no single parameter was generally accepted as a good marker. Recently several studies suggested that adhesion molecules, especially ICAM-1 can be a marker, but there are some controversies. And only few data are available about the relationship of ICAM-1 with clinical follow-up course. Methods: We measured the expression of adhesion molecules on BAL cells by flow cytometry and the level of soluble ICAM-1(sICAM-1) in serum and BALF at the time of diagnosis in 12 patients with active disease and 7 inactive sarcoidosis(5 male, 14 female, mean age: $39.4{\pm}10.7$ years, mean follow-up : $20{\pm}15$ months). Follow-up clinical course were compared with the changes in serum sICAMA-1 level and the adhesion molecule on BAL cells. Results: In the patients with active disease, the ICAM-1 on AM(RMFI: $3.68{\pm}1.71$) and sICAM-1 level in serum($582{\pm}193$ng/ml) and BAL fluid($47.8{\pm}16.5$ng/ml) were all higher than those of 7 inactive disease(RMFI: $1.89{\pm}0.75$, p=0.0298, serum: $294{\pm}117$ ng/ml, p=0.0049, BALF: $20.9{\pm}8.3$ ng/ml). In the active sarcoidosis, ICAM-1 on AM(RMFI : $1.51{\pm}0.84$) and serum sICAM-1 were decreased after the therapy($250{\pm}147$ ng/ml) but no significant change was noted in inactive disease. Also we found the initial ICAM-1 on AM and serum sICAM-1 had a significant correlation with the degree of improvement in PFT after the therapy. During the follow-up, the disease relapsed in 4 patients after the discontinuation of steroid and the serum sICAM-1 level went-up again at the time of relapse. Conclusion: Our data suggest that the serum sICAM-1 level and the ICAM-1 expression on AM can be a good marker of disease activity and also a predictor of outcome in sarcoidosis.

• Korean journal of applied entomology
• /
• v.25 no.4 s.69
• /
• pp.221-228
• /
• 1986

An empirical model to predict initial disease occurrence and subsequent progress of Alternaria leaf spot was constructed based on the modified degree day temperature and frequency of rainfall in three years field experiments. Climatic factors were analized 10-day bases, beginning April 20 to the end of August, and were used as variables for model construction. Cumulative degree portion (CDP) that is over $10^{\circ}C$ in the daily average temperature was used as a parameter to determine the relationship between temperature and initial disease occurrence. Around one hundred and sixty of CDP was needed to initiate disease incidence. This value was considered as temperature threshhold. After reaching 160 CDP, time of initial occurrence was determined by frequency of rainfall. At least four times of rainfall were necessary to be accumulated for initial occurrence of the disease after passing temperature threshhold. Disease progress after initial incidence generally followed the pattern of frequency of rainfall accumulated in those periods. Apparent infection rate (r) in the general differential equation dx/dt=xr(1-x) for individual epidemics when x is disease proportion and t is time, was a linear function of accumulation rate of rainfall frequency (Rc) and was able to be directly estimated based on the equation r=1.06Rc-0.11($R^2=0.993$). Disease severity (x) after t time could be predicted using exponential equation $[x/(1-x)]=[x_0/(1-x)]e^{(b_0+b_1R_c)t}$ derived from the differential equation, when $x_0$ is initial disease, $b_0\;and\;b_1$ are constants. There was a significant linear relationship between disease progress and cumulative number of air-borne conidia of Alternaria mali. When the cumulative number of air-borne conidia was used as an independent variable to predict disease severity, accuracy of prediction was poor with $R^2=0.3328$.

• Progress in Medical Physics
• /
• v.15 no.4
• /
• pp.202-209
• /
• 2004

Flat panel based digital radiography (DR) systems have recently become useful and important in the field of diagnostic radiology. For DRs with amorphous silicon photosensors, CsI(TI) is normally used as the scintillator, which produces visible light corresponding to the absorbed radiation energy. The visible light photons are converted into electric signal in the amorphous silicon photodiodes which constitute a two dimensional array. In order to produce good quality images, detailed behaviors of DR detectors to radiation must be studied. The relationship between air exposure and the DR outputs has been investigated in many studies. But this relationship was investigated under the condition of the fixed tube voltage. In this study, we investigated the relationship between the DR outputs and X-ray in terms of the absorbed energy in the detector rather than the air exposure using SPEC-l8, an X-ray energy spectrum model. Measured exposure was compared with calculated exposure for obtaining the inherent filtration that is a important input variable of SPEC-l8. The absorbed energy in the detector was calculated using algorithm of calculating the absorbed energy in the material and pixel values of real images under various conditions was obtained. The characteristic curve was obtained using the relationship of two parameter and the results were verified using phantoms made of water and aluminum. The pixel values of the phantom image were estimated and compared with the characteristic curve under various conditions. It was found that the relationship between the DR outputs and the absorbed energy in the detector was almost linear. In a experiment using the phantoms, the estimated pixel values agreed with the characteristic curve, although the effect of scattered photons introduced some errors. However, effect of a scattered X-ray must be studied because it was not included in the calculation algorithm. The result of this study can provide useful information about a pre-processing of digital radiography.

• Korean Journal of Environmental Biology
• /
• v.36 no.4
• /
• pp.479-487
• /
• 2018

The purpose of this study is to analyze photochemical activity of nursery seedlings under drought stress, using chlorophyll fluorescence reaction analysis. Young nursery seedlings of tomato (Lycopersicon esculentum Mill.) and cucumber (Cucumis sativa L.), were grown under drought stress for 8 days. Analysis of chlorophyll fluorescence reaction (OJIP) and parameters, were performed to evaluate photochemical fluctuation in nursery seedlings under drought stress. Chlorophyll fluorescence reaction analysis showed maximal recorded fluorescence (P) decreased from the 5 day after treatment in tomato seedlings, while an amount of chlorophyll fluorescence increased at the J-I step. Thus, physiological activity was reduced. In cucumber seedlings, maximal recorded fluorescence (P) and maximal variable fluorescence ($F_V$) lowered from the 4 day after treatment, and chlorophyll fluorescence intensity of J-I step increased. Chlorophyll fluorescence parameter analysis showed electron transfer efficiency of PSII and PSI were significantly inhibited with decreasing $ET2_O/RC$ and $RE1_O/RC$ from the 5 day after treatment, in tomato seedlings and from the 4 day after treatment, in cucumber seedlings. $ET2_O/RC$ and $PI_{ABS}$ significantly changed. In conclusion, 6 indices such as $F_V/F_M$, $DI_O/RC$, $ET2_O/RC$, $RE1_O/RC$, $PI_{ABS}$ and $PI_{TOTAL}ABS$ were selected for determining drought stress in nursery seedlings. Drought stress factor index (DFI) was used to evaluate whether the crop was healthy or not, under drought stress. Cucumber seedlings were less resistant to drought stress than tomato seedlings, in the process of drought stress.

• Tuberculosis and Respiratory Diseases
• /
• v.46 no.2
• /
• pp.215-228
• /
• 1999

Background: Sarcoidosis is a chronic granulomatous inflammatory disease of unknown etiology often involving the lungs and intrathoracic lymph nodes. The natural course of sarcoidosis is variable from spontaneous remission to significant morbidity or death. But, the mechanisms causing the variable clinical outcomes or any single parameter to predict the prognosis was not known. In sarcoidosis, the number and the activity of CD4 + lymphocytes are significantly increased at the loci of disease and their oligoclonality suggests that the CD4 + lymphocytes hyperreactivity may be caused by persistent antigenic stimulus. Recently, it has been known that CD4+ lymphocytes can be subdivided into 2 distinct population(Th1 and Th2) defined by the spectrum of cytokines produced by these cells. Th1 cells promote cellular immunity associated with delayed type hypersensitivity reactions by generating IL-2 and IFN-$\gamma$. Th2 cells playa role in allergic responses and immediate hypersensitivity reactions by secreting IL-4, IL-5, and IL-10. CD4+ lymphocytes in pulmonary sarcoidosis were reported to be mainly Th1 cells. IL-12 has been known to play an important role in differentiation of undifferentiated naive T cells to Th1 cells. And, Moller et al. observed increased IL-12 in bronchoalveolar lavage fluid(BALF) in patients with sarcoidosis. So it is possible that the elevated level of IL-12 is necessary for the continuous progression of the disease in active sarcoidosis. This study was performed to test the assumption that IL-12 can be a marker of active pulmonary sarcoidosis. Methods: We measured the concentration of IL-12 in BALF and in conditioned medium of alveolar macrophage(AM) using ELISA(enzyme-linked immunosorbent assay) method in 26 patients with pulmonary sarcoidosis(10 males, 16 females, mean age: $39.8{\pm}2.1$ years) and 11 normal control. Clinically, 14 patients had active sarcoidosis and 12 patients had inactive. Results: Total cells counts, percentage and number of lymhocytes, number of AM and CD4/CD8 lymphocyte ratio in BALF were significantly higher in patients with sarcoidosis than in control group. But none of these parameters could differentiate active sarcoidosis from inactive disease. The concentration of IL-12 in BALF was significantly increased in sarcoidosis patients ($49.3{\pm}9.2$ pg/ml) than in normal control ($2.5{\pm}0.4$ pg/ml) (p<0.001). Moreover it was significantly higher in patients with active sarcoidosis ($70.3{\pm}14.8$ pg/ml) than in inactive disease ($24.8{\pm}3.l$ pg/ml) (p=0.001). Also, the concentration of IL-12 in BALF showed significant correlation with the percentage of AM(p<0.001), percentage(p<0.001) and number of lymphocyte(p<0.001) in BALF, suggesting the close relationship between the level of IL-12 in BALF and the inflammatory cell infiltration in the lungs. Furthermore, we found a significant correlation between the level of IL-12 and the concentration of soluble ICAM-1 : in serum(p<0.001) and BALF (p=0.001), and also between IL-12 level and ICAM-1 expression of AM(p<0.001). The AM from patients with pulmonary sarcoidosis secreted significantly larger amount of IL-12 ($206.2{\pm}61.9$ pg/ml) than those of control ($68.3{\pm}43.7$ pg/ml) (p<0.008), but, there was no difference between inactive and active disease group. Conclusion : Our data suggest that the BALF IL-12 level can be used as a marker of the activity of pulmonary sarcoidosis.

• Journal of Global Scholars of Marketing Science
• /
• v.18 no.3
• /
• pp.115-132
• /
• 2008

Metal industry is one of the most representative heavy industries and the median sales volume of steel and nonferrous metal companies is over one billion dollars in the case America [Forbes 2006]. As seen in the recent business market situation, an increasing number of industrial manufacturers and suppliers are moving from adversarial to cooperative exchange attitudes that support the long-term relationships with their customers. This article presents the results of an empirical study of the antecedent factors of business relationships in metal industry of the United States. Commitment has been reviewed as a significant and critical variable in research on inter-organizational relationships (Hong et al. 2007, Kim et al. 2007). The future stability of any buyer-seller relationship depends upon the commitment made by the interactants to their relationship. Commitment, according to Dwyer et al. [1987], refers to "an implicit or explicit pledge of relational continuity between exchange partners" and they consider commitment to be the most advanced phase of buyer-seller exchange relationship. Bonds are made because the members need their partners in order to do something and this integration on a task basis can be either symbiotic or cooperative (Svensson 2008). To the extent that members seek the same or mutually supporting ends, there will be strong bonds among them. In other words, the principle that affects the strength of bonds is 'economy of decision making' [Turner 1970]. These bonds provide an important idea to study the causes of business long-term relationships in a sense that organizations can be mutually bonded by a common interest in the economic matters. Recently, the framework of structural bonding has been used to study the buyer-seller relationships in industrial marketing [Han and Sung 2008, Williams et al. 1998, Wilson 1995] in that this structural bonding is a crucial part of the theoretical justification for distinguishing discrete transactions from ongoing long-term relationships. The major antecedent factors of buyer commitment such as technology, CLalt, transaction-specific assets, and importance were identified and explored from the perspective of structural bonding. Research hypotheses were developed and tested by using survey data from the middle managers in the metal industry. H1: Level of technology of the relationship partner is positively related to the level of structural bonding between the buyer and the seller. H2: Comparison level of alternatives is negatively related to the level of structural bonding between the buyer and the seller. H3: Amount of the transaction-specific assets is positively related to the level of structural bonding between the buyer and the seller. H4: Importance of the relationship partner is positively related to the level of structural bonding between the buyer and the seller. H5: Level of structural bonding is positively related to the level of commitment to the relationship. To examine the major antecedent factors of industrial buyer's structural bonding and long-term relationship, questionnaire was prepared, mailed out to the sample of 400 purchasing managers of the US metal industry (SIC codes 33 and 34). After a follow-up request, 139 informants returnedthe questionnaires, resulting in a response rate of 35 percent. 134 responses were used in the final analysis after dropping 5 incomplete questionnaires. All measures were analyzed for reliability and validity following the guidelines offered by Churchill [1979] and Anderson and Gerbing [1988]., the results of fitting the model to the data indicated that the hypothesized model provides a good fit to the data. Goodness-of-fit index (GFI = 0.94) and other indices ( chi-square = 78.02 with p-value = 0.13, Adjusted GFI = 0.90, Normed Fit Index = 0.92) indicated that a major proportion of variances and covariances in the data was accounted for by the model as a whole, and all the parameter estimates showed statistical significance as evidenced by large t-values. All the factor loadings were significantly different from zero. On these grounds we judged the hypothesized model to be a reasonable representation of the data. The results from the present study suggest several implications for buyer-seller relationships. Theoretically, we attempted to conceptualize the antecedent factors of buyer-seller long-term relationships from the perspective of structural bondingin metal industry. The four underlying determinants (i.e. technology, CLalt, transaction-specific assets, and importance) of structural bonding are very critical variables of buyer-seller long-term business relationships. Our model of structural bonding makes an attempt to systematically examine the relationship between the antecedent factors of structural bonding and long-term commitment. Managerially, this research provides industrial purchasing managers with a good framework to assess the interaction processes with their partners and, ability to position their business relationships from the perspective of structural bonding. In other words, based on those underlying variables, industrial purchasing managers can determine the strength of the company's relationships with the key suppliers and its state of preparation to be a successful partner with those suppliers. Both the supplying and customer companies can also benefit by using the concept of 'structural bonding' and evaluating their relationships with key business partners from the structural point of view. In general, the results indicate that structural bonding gives a critical impact on the level of relationship commitment. Managerial implications and limitations of the study are also discussed.

• Information Systems Review
• /
• v.16 no.3
• /
• pp.161-177
• /
• 2014

Corporate credit rating assessment consists of complicated processes in which various factors describing a company are taken into consideration. Such assessment is known to be very expensive since domain experts should be employed to assess the ratings. As a result, the data-driven corporate credit rating prediction using statistical and artificial intelligence (AI) techniques has received considerable attention from researchers and practitioners. In particular, statistical methods such as multiple discriminant analysis (MDA) and multinomial logistic regression analysis (MLOGIT), and AI methods including case-based reasoning (CBR), artificial neural network (ANN), and multiclass support vector machine (MSVM) have been applied to corporate credit rating.2) Among them, MSVM has recently become popular because of its robustness and high prediction accuracy. In this study, we propose a novel optimized MSVM model, and appy it to corporate credit rating prediction in order to enhance the accuracy. Our model, named 'GAMSVM (Genetic Algorithm-optimized Multiclass Support Vector Machine),' is designed to simultaneously optimize the kernel parameters and the feature subset selection. Prior studies like Lorena and de Carvalho (2008), and Chatterjee (2013) show that proper kernel parameters may improve the performance of MSVMs. Also, the results from the studies such as Shieh and Yang (2008) and Chatterjee (2013) imply that appropriate feature selection may lead to higher prediction accuracy. Based on these prior studies, we propose to apply GAMSVM to corporate credit rating prediction. As a tool for optimizing the kernel parameters and the feature subset selection, we suggest genetic algorithm (GA). GA is known as an efficient and effective search method that attempts to simulate the biological evolution phenomenon. By applying genetic operations such as selection, crossover, and mutation, it is designed to gradually improve the search results. Especially, mutation operator prevents GA from falling into the local optima, thus we can find the globally optimal or near-optimal solution using it. GA has popularly been applied to search optimal parameters or feature subset selections of AI techniques including MSVM. With these reasons, we also adopt GA as an optimization tool. To empirically validate the usefulness of GAMSVM, we applied it to a real-world case of credit rating in Korea. Our application is in bond rating, which is the most frequently studied area of credit rating for specific debt issues or other financial obligations. The experimental dataset was collected from a large credit rating company in South Korea. It contained 39 financial ratios of 1,295 companies in the manufacturing industry, and their credit ratings. Using various statistical methods including the one-way ANOVA and the stepwise MDA, we selected 14 financial ratios as the candidate independent variables. The dependent variable, i.e. credit rating, was labeled as four classes: 1(A1); 2(A2); 3(A3); 4(B and C). 80 percent of total data for each class was used for training, and remaining 20 percent was used for validation. And, to overcome small sample size, we applied five-fold cross validation to our dataset. In order to examine the competitiveness of the proposed model, we also experimented several comparative models including MDA, MLOGIT, CBR, ANN and MSVM. In case of MSVM, we adopted One-Against-One (OAO) and DAGSVM (Directed Acyclic Graph SVM) approaches because they are known to be the most accurate approaches among various MSVM approaches. GAMSVM was implemented using LIBSVM-an open-source software, and Evolver 5.5-a commercial software enables GA. Other comparative models were experimented using various statistical and AI packages such as SPSS for Windows, Neuroshell, and Microsoft Excel VBA (Visual Basic for Applications). Experimental results showed that the proposed model-GAMSVM-outperformed all the competitive models. In addition, the model was found to use less independent variables, but to show higher accuracy. In our experiments, five variables such as X7 (total debt), X9 (sales per employee), X13 (years after founded), X15 (accumulated earning to total asset), and X39 (the index related to the cash flows from operating activity) were found to be the most important factors in predicting the corporate credit ratings. However, the values of the finally selected kernel parameters were found to be almost same among the data subsets. To examine whether the predictive performance of GAMSVM was significantly greater than those of other models, we used the McNemar test. As a result, we found that GAMSVM was better than MDA, MLOGIT, CBR, and ANN at the 1% significance level, and better than OAO and DAGSVM at the 5% significance level.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.23 no.3
• /
• pp.225-237
• /
• 1990

Various chemical constituents were measured from April to August 1988 at the down-ward 20 stations of Keum River, which is located in the Midwest of Korea, to understand the characteristics of water quality with respect to spatio-temporal variations of each constituent. The 24-hrs continuous measurements with 2-hrs interval were made simultaneously at station 2 near the estuary weir and station 9(Ganggyeong) of 35 km upstream from the weir in April. By the results observed for one day in April at station 2, salinity has a range of $7.88\~22.14\%_{\circ}$ and its temporal variability is identical to the pattern of tidal cycle in the neigh-bouring Kunsan Harbor. However, turbidity shows relatively high values only at an interval of 4~5 hours after the lowest salinity time, though hourly fluctuation of pH is very small. Silicate and dissolved inorganic nitrogen have inversively linear correlationships with salinity, implying the concentration of the two nutrients strongly regulated by estuarine mixing of sea and river waters. In contrast, phosphate sustains roughly a constant level over a wide salinity range and distinctly lower values than those corresponding to nitrate in the oceans. Such distributions of phosphate have been observed in some estuaries, and interpreted as driven by removal of dissolved phosphate into bottom sediments and the bufforing of phosphate by particulate matter. COD values at station 2 are relatively high in day-time(particularly afternoon) and in high-salinity periods. At station 9, saltwater intrusion was never found but water level changed to the extent of 2.5 m for one day. Although each parameter at this station exhibits very slight variations in their abundance for 24 hours compared with station 2, the contents of COD, silicate and ammonia are significantly higher than at station 2. Concentration of suspended matter is relatively high in the brackish water region up to $\~20$ km above the river mouth, probably due to strong tidal stirring of the bottom de-posits. Also, relatively high pH, COD and $O_2$ saturation at the upward stations of $40\~50$ km from the weir are presumably attributable to active photosynthesis of plants in the region. In general, COD and nutrients except phosphate are higher values at the upper stations than in the estuary zone, and show the highest abundances in July nearly at all stations. Finally, in the estuarine region tidal mixing of sea-river waters seems to be an important factor controlling the distributions of turbidity, COD, silicate and nitrate as well as salinity. However, water quality in the upward fresh-water zone is remarkably variable according to months or seasons.