• Journal of Intelligence and Information Systems
• /
• v.24 no.2
• /
• pp.221-241
• /
• 2018

Deep learning is getting attention recently. The deep learning technique which had been applied in competitions of the International Conference on Image Recognition Technology(ILSVR) and AlphaGo is Convolution Neural Network(CNN). CNN is characterized in that the input image is divided into small sections to recognize the partial features and combine them to recognize as a whole. Deep learning technologies are expected to bring a lot of changes in our lives, but until now, its applications have been limited to image recognition and natural language processing. The use of deep learning techniques for business problems is still an early research stage. If their performance is proved, they can be applied to traditional business problems such as future marketing response prediction, fraud transaction detection, bankruptcy prediction, and so on. So, it is a very meaningful experiment to diagnose the possibility of solving business problems using deep learning technologies based on the case of online shopping companies which have big data, are relatively easy to identify customer behavior and has high utilization values. Especially, in online shopping companies, the competition environment is rapidly changing and becoming more intense. Therefore, analysis of customer behavior for maximizing profit is becoming more and more important for online shopping companies. In this study, we propose 'CNN model of Heterogeneous Information Integration' using CNN as a way to improve the predictive power of customer behavior in online shopping enterprises. In order to propose a model that optimizes the performance, which is a model that learns from the convolution neural network of the multi-layer perceptron structure by combining structured and unstructured information, this model uses 'heterogeneous information integration', 'unstructured information vector conversion', 'multi-layer perceptron design', and evaluate the performance of each architecture, and confirm the proposed model based on the results. In addition, the target variables for predicting customer behavior are defined as six binary classification problems: re-purchaser, churn, frequent shopper, frequent refund shopper, high amount shopper, high discount shopper. In order to verify the usefulness of the proposed model, we conducted experiments using actual data of domestic specific online shopping company. This experiment uses actual transactions, customers, and VOC data of specific online shopping company in Korea. Data extraction criteria are defined for 47,947 customers who registered at least one VOC in January 2011 (1 month). The customer profiles of these customers, as well as a total of 19 months of trading data from September 2010 to March 2012, and VOCs posted for a month are used. The experiment of this study is divided into two stages. In the first step, we evaluate three architectures that affect the performance of the proposed model and select optimal parameters. We evaluate the performance with the proposed model. Experimental results show that the proposed model, which combines both structured and unstructured information, is superior compared to NBC(Naïve Bayes classification), SVM(Support vector machine), and ANN(Artificial neural network). Therefore, it is significant that the use of unstructured information contributes to predict customer behavior, and that CNN can be applied to solve business problems as well as image recognition and natural language processing problems. It can be confirmed through experiments that CNN is more effective in understanding and interpreting the meaning of context in text VOC data. And it is significant that the empirical research based on the actual data of the e-commerce company can extract very meaningful information from the VOC data written in the text format directly by the customer in the prediction of the customer behavior. Finally, through various experiments, it is possible to say that the proposed model provides useful information for the future research related to the parameter selection and its performance.

• Journal of Hospice and Palliative Care
• /
• v.5 no.2
• /
• pp.111-124
• /
• 2002

Purpose : Although the average life expectancy has increased due to advances in medicine, mortality due to cancer is on an increasing trend. Consequently, the number of terminally ill cancer patients is also on the rise. Predicting the survival period is an important issue in the treatment of terminally ill cancer patients since the choice of treatment would vary significantly by the patents, their families, and physicians according to the expected survival. Therefore, we investigated the prognostic factors for increased mortality risk in terminally ill cancer patients to help treat these patients by predicting the survival period. Methods : We investigated 31 clinical parameters in 157 terminally ill cancer patients admitted to in the Department of Family Medicine, National Health Insurance Corporation Ilsan Hospital between July 1, 2000 and August 31, 2001. We confirmed the patients' survival as of October 31, 2001 based on medical records and personal data. The survival rates and median survival times were estimated by the Kaplan-Meier method and Log-rank test was used to compare the differences between the survival rates according to each clinical parameter. Cox's proportional hazard model was used to determine the most predictive subset from the prognostic factors among many clinical parameters which affect the risk of death. We predicted the mean, median, the first quartile value and third quartile value of the expected lifetimes by Weibull proportional hazard regression model. Results : Out of 157 patients, 79 were male (50.3%). The mean age was $65.1{\pm}13.0$ years in males and was $64.3{\pm}13.7$ years in females. The most prevalent cancer was gastric cancer (36 patients, 22.9%), followed by lung cancer (27, 17.2%), and cervical cancer (20, 12.7%). The survival time decreased with to the following factors; mental change, anorexia, hypotension, poor performance status, leukocytosis, neutrophilia, elevated serum creatinine level, hypoalbuminemia, hyperbilirubinemia, elevated SGPT, prolonged prothrombin time (PT), prolonged activated partial thromboplastin time (aPTT), hyponatremia, and hyperkalemia. Among these factors, poor performance status, neutrophilia, prolonged PT and aPTT were significant prognostic factors of death risk in these patients according to the results of Cox's proportional hazard model. We predicted that the median life expectancy was 3.0 days when all of the above 4 factors were present, $5.7{\sim}8.2$ days when 3 of these 4 factors were present, $11.4{\sim}20.0$ days when 2 of the 4 were present, and $27.9{\sim}40.0$ when 1 of the 4 was present, and 77 days when none of these 4 factors were present. Conclusions : In terminally ill cancer patients, we found that the prognostic factors related to reduced survival time were poor performance status, neutrophilia, prolonged PT and prolonged am. The four prognostic factors enabled the prediction of life expectancy in terminally ill cancer patients.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.14 no.1
• /
• pp.94-100
• /
• 2010

Purpose: We can evaluate function of kidney by Glomerular Filtration Rate (GFR) test using $^{99m}Tc$-DTPA which is simple. This test is influenced by several parameter such as net syringe count, kidney depth, corrected kidney count, acquisition time and characters of gamma camera. In this study, we evaluated predose count according to matrix size in the GFR test using $^{99m}Tc$-DTPA. Materials and Methods: Gamma camera of Infinia in GE was used, and LEGP collimator, three types of matrix size ($64{\times}64$, $128{\times}128$, $256{\times}256$) and 1.0 of zoom factor were applied. We increased radioactivity concentration from 222 (6), 296 (8), 370 (10), 444 (12) up to 518 MBq (14 mCi) respectively and acquired images according to matrix size at 30 cm distance from detector. Lastly, we evaluated these values and then substituted them for GFR formula. Results: In $64{\times}64$, $128{\times}128$ and $256{\times}256$ of matrix size, counts per second was 26.8, 34.5, 41.5, 49.1 and 55.3 kcps, 25.3, 33.4, 41.0, 48.4 and 54.3 kcps and 25.5, 33.7, 40.8, 48.1 and 54.7 kcps respectively. Total counts for 5 second were 134, 172, 208, 245 and 276 kcounts from $64{\times}64$, 127, 172, 205, 242, 271 kcounts from $128{\times}128$, and 137, 168, 204, 240 and 273 kcounts from $256{\times}256$, and total counts for 60 seconds were 1,503, 1,866, 2,093, 2,280, 2,321 kcounts, 1,511, 1,994, 2,453, 2,890 and 3,244 kcounts, and 1,524, 2,011, 2,439, 2,869 and 3,268 kcounts respectively. It is different from 0 to 30.02 % of percentage difference in $64{\times}64$ of matrix size. But in $128{\times}128$ and $256{\times}256$, it is showed 0.60 and 0.69 % of maximum value each. GFR of percentage difference in $64{\times}64$ represented 6.77% of 222 MBq (6 mCi), 42.89 % of 518 MBq (14 mCi) at 60 seconds respectively. However it is represented 0.60 and 0.63 % each in $128{\times}128$ and $256{\times}256$. Conclusion: There was no big difference in total counts of percentage difference and GFR values acquiring from $128{\times}128$ and $256{\times}256$ of matrix size. But in $64{\times}64$ of matrix size when the total count exceeded 1,500 kcounts, the overflow phenomenon was appeared differently according to predose radioactivity of concentration and acquisition time. Therefore, we must optimize matrix size and net syringe count considering the total count of predose to get accurate GFR results.

• Journal of Chest Surgery
• /
• v.39 no.9 s.266
• /
• pp.692-698
• /
• 2006

Background: Owing to the fact that the average life span has increased and the progress in medical science has been made, the number of patients with chronic renal failure (CRF) who have to take hemodialysis (HD) has been going up gradually. Accordingly, it is considered to be as a significant issue to obtain blood vessels which can be used repetitively and supply enough blood flows. Therefore, there have been various kinds of study on an inosculation rate andfactors influencing it following an arteriovenous fistula (AV fistula) and lots of studies are ongoing for the purpose of escalating the inosculation rate. The authors analyzed the effects of short-term result, age, sex, diabetes and hypertension on arteriovenous inosculations in 134 anatomical snuffbox operated subjects among the patients who have taken an AV fistula at this center. Material and Method: Based on 134 patients who underwent an AV fistula at the department of thoracic surgery of this center from July, 2000 to May, 2004, the difference in arteriovenous inosculation rate was compared and analyzed depending or age (discriminated by 65-year-old), sex and the condition of the presence or absence of diabetes and hypertension. Correlation analyses were conducted for each parameter and statistical tests were performed by using SPSS for windows Release 11.0.1, which were determined to be statistically significant if p value was below 0.05. Result: The total number of operations was 169 including 35 of re-operations. The male/female rate was 70 : 64 (52% : 48%). The average age was $56.3{\pm}12.26$ years and there were 33 (24%) old aged patients above 65-year-old; there were 103 (71%) patients with hypertension and 90 (67%) patients with diabetes. Overall arteriovenous inosculation rate was $93{\pm}2.4%,\;91{\pm}2.7%,\;89{\pm}3.0%$ at 6, 12, 24 months, respectively. The arteriovenous inosculation rate of above 65-year-old patient group was $85{\pm}4.8%,\;80{\pm}5.8%,\;80{\pm}5.8%$ and below 64-year-old patient group's was $85{\pm}4.8%,\;80{\pm}5.8%,\;80{\pm}5.8%$ at given time points, respectively, which showed higher inosculation rate in below 64-year-old patient group with a statistical significance (p=0.0034). However, no statistical significance was found between the patients with hypertension and diabetes and the patients with no complication. In addition, there was no statistical significance in inosculation rate between male and female. Conclusion: The arteriovenous inosculation ratewas higher in the treated patient below 64-year-old than in the treated patient above 65-year-old. Thus it is advantageous for increase in long-term inosculation rate to obtain hemodialysis routes at an early age. The conditions of sex and the presence or absence of diabetes and hyper- tension do not make statistically significant effect on the arteriovenous inosculation rate.

• The Journal of Small Business Innovation
• /
• v.19 no.4
• /
• pp.19-39
• /
• 2016

With increasing adoption of smart products and complexity, companies have shifted their strategies from stand alone and competitive strategies to business ecosystem oriented and cooperative strategies. The win-win growth of business refers to corporate efforts undertaken by companies to pursue the healthiness of business between conglomerates and partnering companies such as suppliers for mutual prosperity and a long-term corporate soundness based on their business ecosystem and cooperative strategies. This study is designed to validate a theoretical proposition that the win-win growth strategy of Samsung Electronics and cooperative efforts among companies can create a healthy business ecosystem, based on results of case studies and surveys. In this study, a level of global market access of small and mid-sized companies is adopted as the key achievement index. The foreign market entry is considered as one of vulnerabilities in the ecosystem of small and mid-sized enterprises (SMEs). For SMEs, the global market access based on the research and development (R&D) has become the critical component in the process of transforming them into global small giants. The results of case studies and surveys are analyzed mainly based on a model of a virtuous cycle of Creativity, Opportunity, Productivity, and Proactivity (the COPP model) that features the characteristics of the healthiness of a business ecosystem. In the COPP model, a virtuous circle of profits made by the first three factors and Proactivity, which is the manifestation of entrepreneurship that proactively invests and reacts to the changing business environment of the future, enhances the healthiness of a given business ecosystem. With the application of the COPP model, this study finds major achievements of the win-win growth of Samsung Electronics as follows. First, Opportunity plays a role as a parameter in the relations of Creativity, Productivity, and creating profits. Namely, as companies export more (with more Opportunity), they are more likely to link their R&D efforts to Productivity and profitability. However, companies that do not export tend to fail to link their R&D investment to profitability. Second, this study finds that companies with huge investment on R&D for the future, which is the result of Proactivity, tend to hold a large number of patents (Creativity). And companies with significant numbers of patents tend to be large exporters as well (Opportunity), and companies with a large amount of exports tend to record high profitability (Productivity and profitability), and thus forms the virtuous cycle of the COPP model. In addition, to access global markets for sustainable growth, SMEs need to build and strengthen their competitiveness. This study concludes that companies with a high level of proactivity to invest for the future can create a virtuous circle of Creativity, Opportunity, Productivity, and Proactivity, thereby providing a strategic implication that SMEs should invest time and resources in forming such a virtuous cycle which is a sure way for the SMEs to grow into global small giants.

• 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.

• Journal of Animal Science and Technology
• /
• v.52 no.5
• /
• pp.349-356
• /
• 2010

This study was conducted to estimate heritabilities, repeatabilities and rank correlation coefficients among breeding values for litter size and sex ratio of Yorkshire and Landrace pigs using various single trait animal models. The analyses were carried out the data comprising 26,390 litters of Yorkshire and 26,173 litters of Landrace collected from the year 1998 to 2008 at a private swine breeding farm located in central part of Korea. Five different analytical models were used for genetic parameter estimation. Model 1 was most simple basic model fitted with year-month contemporary group fixed effect, random additive genetic effect and random residual effect. Model 2 was similar to the model 1 but permanent maternal environmental effect added as random effect, and model 3 was similar with the model 2 but linear and quadratic effects of sow age were added as fixed covariate effect. Model 4 was similar as model 2 except that the parity was added as fixed effect and model 5 was similar to model 3 or model 4 but covariate of sow age was nested within parity effect. The results obtained in this study are summarized as follows: The means and standard error of total number of pigs born per litter (TNB) and number of pigs born alive per litter (NBA) were $11.35{\pm}0.02$ and $10.04{\pm}0.02$ for Yorkshire, $10.97{\pm}0.02$ and $9.98{\pm}0.02$ for Landrace, respectively. The sex ratio (percentage of female per litter) was $45.75{\pm}0.11%$ and $45.75{\pm}0.11%$ for Yorkshire and Landrace, respectively. The heritability estimates of TNB (0.243) and NBA (0.192) from model 1 tended to be higher than those from any other models in both breeds. Differences in heritability and repeatability for TNB were not large among models 3, 4 and 5 and same tendency of negligible differences among estimates by models 3, 4 and 5 were observed for NBA, where heritability and repeatability ranged from 0.096 to 0.099 and from 0.188 to 0.193, respectively, in Yorkshire; and ranged from 0.092 to 0.098 and from 0.193 and 0.196, respectively, in Landrace. The heritability estimates for sex ratio were close to zero which was ranged from 0.002 to 0.003 for TNB and from 0.001 to 0.003 for NBA over the models applied. The rank correlation coefficients of breeding values by model 1 with those from other models (model 2, 3, 4 and 5), and breeding values by model 2 with those from other models (model 1, 3, 4 and 5) were highly positive but lower than the coefficients among breeding values by model 3, model 4 and model 5 which were high of 0.99, approximately, for TNB and NBA of both breeds.

• Korean Journal of Weed Science
• /
• v.9 no.1
• /
• pp.56-68
• /
• 1989

aA series of experiments were conducted to investigate the effect of plant growth regulator treatments on the growth and lateral root formation in soybean sprouts in order to establish the effective method of producing root-less or short-rooted soybean sprouts with larger diameter in the hypocotyl. Major results can be summarized as follows. 1. Soybean sprouts showed fairly uniform elongation rate from 3 to g days after imbibition with daily increase of 3.8cm. The speed of elongation of hypocotyl was reduced whereas that of root accelerated 7 days after imbibition. Lateral roots began to emerge fairly evenly from 5 to 9 days after imbibition with a daily increase of 4.4. 2. Auxins(IAA, IBA, NAA, 2,4-D) inhibited hypocotyl elongation and formation of lateral roots and increased hypocotyl diameter without influencing root length and hook diameter at higher concentrations. The dry weight of cotyledon was increased significantly as compared to that of hypocotyl and root. Among the tested auxins, 2, 4-D was the most effective. 3. BA and 4PU-30 significantly reduced elongation of hypocotyl and root and resulted in the biggest diameter of hypocotyl when treated at higher concentrations. The lowest effective concentration of BA to prevent the formation of larval gal roots was 12.5ppm. The formation of lateral roots could be completely prevented by BA and 4PU-30 treatment but kinetin, zeatin, zeatin riboside resulted in many lateral roots and increased thickness of soybean sprouts with little influence. Cotyledon deformation was found in soybean sprouts treated by 4PU-30. 4. 2, 4-D was the most effective for increasing the hypocotyl diameter while 4PU-30 was the most effective for reducing no. of lateral roots. 5. It can be concluded that among the plant growth regulators tested, BA was effective in reducing root length and increasing hypocotyl diameter. BA 12.5 ppm or 15 ppm may thus be the more practical for production of soybean sprouts. 6. ABA showed no significant effect of growth parameter, however ABA 25 ppm inhibited only no of lateral roots with little influence on the growth of seedling. 7. Ethephon inhibited the elongation of hypocotyl and root and increased hypocotyl diameter at higher concentrations. 8. The combined effect of cytokinins and ethephon was very similar to result of BA treatment alone. As the ethephon concentration increased, hypocotyl diameter and dry weight of cotyledon tended to increase.

• The Korean Journal of Nuclear Medicine
• /
• v.31 no.1
• /
• pp.73-82
• /
• 1997

Regional myocardial blood flow (rMBF) can be noninvasively quantified using N-13 ammonia and dynamic positron emission tomography (PET). The quantitative accuracy of the rMBF values, however, is affected by the distortion of myocardial PET images caused by finite PET image resolution and cardiac motion. Although different methods have been developed to correct the distortion typically classified as partial volume effect and spillover, the methods are too complex to employ in a routine clinical environment. We have developed a refined method incorporating a geometric model of the volume representation of a region-of-interest (ROI) into the two-compartment N-13 ammonia model. In the refined model, partial volume effect and spillover are conveniently corrected by an additional parameter in the mathematical model. To examine the accuracy of this approach, studies were performed in 9 coronary artery disease patients. Dynamic transaxial images (16 frames) were acquired with a GE $Advance^{TM}$ PET scanner simultaneous with intravenous injection of 20 mCi N-13 ammonia. rMBF was examined at rest and during pharmacologically (dipyridamole) induced coronary hyperemia. Three sectorial myocardium (septum, anterior wall and lateral wall) and blood pool time-activity curves were generated using dynamic images from manually drawn ROIs. The accuracy of rMBF values estimated by the refined method was examined by comparing to the values estimated using the conventional two-compartment model without partial volume effect correction rMBF values obtained by the refined method linearly correlated with rMBF values obtained by the conventional method (108 myocardial segments, correlation coefficient (r)=0.88). Additionally, underestimated rMBF values by the conventional method due to partial volume effect were corrected by theoretically predicted amount in the refined method (slope(m)=1.57). Spillover fraction estimated by the two methods agreed well (r=1.00, m=0.98). In conclusion, accurate rMBF values can be efficiently quantified by the refined method incorporating myocardium geometric information into the two-compartment model using N-13 ammonia and PET.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.17 no.1
• /
• pp.3642-3654
• /
• 1975

The purpose of this thesis is to derive a unit hydrograph which may be applied to the ungaged watershed area from the relations between directly measurable unitgraph properties such as peak discharge(qp), time to peak discharge (Tp), and lag time (Lg) and watershed characteristics such as river length(L) from the given station to the upstream limits of the watershed area in km, river length from station to centroid of gravity of the watershed area in km (Lca), and main stream slope in meter per km (S). Other procedure based on routing a time-area diagram through catchment storage named Instantaneous Unit Hydrograph(IUH). Dimensionless unitgraph also analysed in brief. The basic data (1969 to 1973) used in these studies are 9 recording level gages and rating curves, 41 rain gages and pluviographs, and 40 observed unitgraphs through the 9 sub watersheds in Nak Oong River basin. The results summarized in these studies are as follows; 1. Time in hour from start of rise to peak rate (Tp) generally occured at the position of 0.3Tb (time base of hydrograph) with some indication of higher values for larger watershed. The base flow is comparelatively higher than the other small watershed area. 2. Te losses from rainfall were divided into initial loss and continuing loss. Initial loss may be defined as that portion of storm rainfall which is intercepted by vegetation, held in deppression storage or infiltrated at a high rate early in the storm and continuing loss is defined as the loss which continues at a constant rate throughout the duration of the storm after the initial loss has been satisfied. Tis continuing loss approximates the nearly constant rate of infiltration (${\Phi}$-index method). The loss rate from this analysis was estimated 50 Per cent to the rainfall excess approximately during the surface runoff occured. 3. Stream slope seems approximate, as is usual, to consider the mainstreamonly, not giving any specific consideration to tributary. It is desirable to develop a single measure of slope that is representative of the who1e stream. The mean slope of channel increment in 1 meter per 200 meters and 1 meter per 1400 meters were defined at Gazang and Jindong respectively. It is considered that the slopes are low slightly in the light of other river studies. Flood concentration rate might slightly be low in the Nak Dong river basin. 4. It found that the watershed lag (Lg, hrs) could be expressed by Lg=0.253 (L.Lca)0.4171 The product L.Lca is a measure of the size and shape of the watershed. For the logarithms, the correlation coefficient for Lg was 0.97 which defined that Lg is closely related with the watershed characteristics, L and Lca. 5. Expression for basin might be expected to take form containing theslope as {{{{ { L}_{g }=0.545 {( { L. { L}_{ca } } over { SQRT {s} } ) }^{0.346 } }}}} For the logarithms, the correlation coefficient for Lg was 0.97 which defined that Lg is closely related with the basin characteristics too. It should be needed to take care of analysis which relating to the mean slopes 6. Peak discharge per unit area of unitgraph for standard duration tr, ㎥/sec/$\textrm{km}^2$, was given by qp=10-0.52-0.0184Lg with a indication of lower values for watershed contrary to the higher lag time. For the logarithms, the correlation coefficient qp was 0.998 which defined high sign ificance. The peak discharge of the unitgraph for an area could therefore be expected to take the from Qp=qp. A(㎥/sec). 7. Using the unitgraph parameter Lg, the base length of the unitgraph, in days, was adopted as {{{{ {T}_{b } =0.73+2.073( { { L}_{g } } over {24 } )}}}} with high significant correlation coefficient, 0.92. The constant of the above equation are fixed by the procedure used to separate base flow from direct runoff. 8. The width W75 of the unitgraph at discharge equal to 75 per cent of the peak discharge, in hours and the width W50 at discharge equal to 50 Per cent of the peak discharge in hours, can be estimated from {{{{ { W}_{75 }= { 1.61} over { { q}_{b } ^{1.05 } } }}}} and {{{{ { W}_{50 }= { 2.5} over { { q}_{b } ^{1.05 } } }}}} respectively. This provides supplementary guide for sketching the unitgraph. 9. Above equations define the three factors necessary to construct the unitgraph for duration tr. For the duration tR, the lag is LgR=Lg+0.2(tR-tr) and this modified lag, LgRis used in qp and Tb It the tr happens to be equal to or close to tR, further assume qpR=qp. 10. Triangular hydrograph is a dimensionless unitgraph prepared from the 40 unitgraphs. The equation is shown as {{{{ { q}_{p } = { K.A.Q} over { { T}_{p } } }}}} or {{{{ { q}_{p } = { 0.21A.Q} over { { T}_{p } } }}}} The constant 0.21 is defined to Nak Dong River basin. 11. The base length of the time-area diagram for the IUH routing is {{{{C=0.9 {( { L. { L}_{ca } } over { SQRT { s} } ) }^{1/3 } }}}}. Correlation coefficient for C was 0.983 which defined a high significance. The base length of the T-AD was set to equal the time from the midpoint of rain fall excess to the point of contraflexure. The constant K, derived in this studies is K=8.32+0.0213 {{{{ { L} over { SQRT { s} } }}}} with correlation coefficient, 0.964. 12. In the light of the results analysed in these studies, average errors in the peak discharge of the Synthetic unitgraph, Triangular unitgraph, and IUH were estimated as 2.2, 7.7 and 6.4 per cent respectively to the peak of observed average unitgraph. Each ordinate of the Synthetic unitgraph was approached closely to the observed one.