• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.60 no.3
• /
• pp.639-647
• /
• 2011

In this paper, we introduce a design methodology of data-centroid Radial Basis Function neural networks with extended polynomial function. The two underlying design mechanisms of such networks involve K-means clustering method and Particle Swarm Optimization(PSO). The proposed algorithm is based on K-means clustering method for efficient processing of data and the optimization of model was carried out using PSO. In this paper, as the connection weight of RBF neural networks, we are able to use four types of polynomials such as simplified, linear, quadratic, and modified quadratic. Using K-means clustering, the center values of Gaussian function as activation function are selected. And the PSO-based RBF neural networks results in a structurally optimized structure and comes with a higher level of flexibility than the one encountered in the conventional RBF neural networks. The PSO-based design procedure being applied at each node of RBF neural networks leads to the selection of preferred parameters with specific local characteristics (such as the number of input variables, a specific set of input variables, and the distribution constant value in activation function) available within the RBF neural networks. To evaluate the performance of the proposed data-centroid RBF neural network with extended polynomial function, the model is experimented with using the nonlinear process data(2-Dimensional synthetic data and Mackey-Glass time series process data) and the Machine Learning dataset(NOx emission process data in gas turbine plant, Automobile Miles per Gallon(MPG) data, and Boston housing data). For the characteristic analysis of the given entire dataset with non-linearity as well as the efficient construction and evaluation of the dynamic network model, the partition of the given entire dataset distinguishes between two cases of Division I(training dataset and testing dataset) and Division II(training dataset, validation dataset, and testing dataset). A comparative analysis shows that the proposed RBF neural networks produces model with higher accuracy as well as more superb predictive capability than other intelligent models presented previously.

• Radiation Oncology Journal
• /
• v.17 no.2
• /
• pp.91-99
• /
• 1999

Propose : This study was performed to find out the prognostic factors affecting local control, survival and disease free survival rate in nasopharyngeal carcinomas treated with chemotherapy and radiation therapy. Materials and Methods : We analysed 47 patients of nasopharyngeal carcinomas, histologically confirmed and treated at Chonnam University Hospital between July 1986 and June 1996, retrospectively. Range of patients' age were from 16 to 80 years (median; 52 years). Thirty three (70$\%$) patients was male. Histological types were composed of 3 (6$\%$) keratinizing, 30 (64$\%$) nonkeratinizing squamous cell carcinoma and 13 (28$\%$) undifferentiated carcinoma. Histoiogicai type was not known in 1 patient (2$\%$). We restaged according to the staging system of 1997 American Joint Committee on Cancer Forty seven patients were recorded as follows: 71: 11 (23$\%$), T2a; 6 (13$\%$), T2b; 9 (19$\%$), 73; 7 (15$\%$), 74: 14 (30$\%$), and NO; 7 (15$\%$), Nl: 14 (30$\%$), N2; 21 (45%), N3: 5 (10%). Clinical staging was grouped as follows: Stage 1; 2 (4$\%$), IIA: 2 (4$\%$), IIB; 10 (21$\%$), III; 14 (30$\%$), IVA; 14 (30$\%$) and IVB; 5 (11$\%$). Radiation therapy was done using 6 MV and 10 MV X- ray of linear accelerator. Electron beam was used for the Iymph nodes of posterior neck after 4500 cGy. The range of total radiation dose delivered to the primary tumor was from 6120 to 7920 cGy (median; 7020 cGy). Neoadjuvant chemotherapy was performed with cisplatin +5-fluorouracil (25 patients) or cisplatin+pepleomycin (17 patients) with one to three cycles. Five patients did not received chemotherapy. Local control rate, survival and disease free suwival rate were calculated by Kaplan-Meier method. Generalized Wilcoxon test was used to evaluate the difference of survival rates between groups. multivariate analysis using Cox proportional hazard model was done for finding prognostic factors. Results: Local control rate was 81$\%$ in 5 year. Five year survival rate was 60$\%$ (median survival; 100 months). We included age, sex, cranial nerve deflicit, histologic type, stage group, chemotherapy, elapsed days between chemotherapy and radiotherapy, total radiation dose, period of radiotherapy as potential prognostic factors in multivariate analysis. As a result, cranial none deficit (P=0.004) had statistical significance in local control rate. Stage group and total radiation dose were significant prognostic factors in survival (P=0.000, P=0.012), and in disease free survival rates (P=0.003, P=0.008), respectively. Common complications were xerostomia, tooth and ear problems. Hypothyroidism was developed in 2 patients. Conclusion : In our study, cranial none deficit was a significant prognostic factor in local control rate, and stage group and total radiation dose were significant factors in both survival and disease free survival of nasopharyngeal carcinoma. We have concluded that chemotherapy and radiotherapy used in our patients were effective without any serious complication.

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.147-170
• /
• 2022

Although the number of patent which is one of the core outputs of technological innovation continues to increase, the number of low-value patents also hugely increased. Therefore, efficient evaluation of patents has become important. Estimation of patent lifespan which represents private value of a patent, has been studied for a long time, but in most cases it relied on a linear model. Even if machine learning methods were used, interpretation or explanation of the relationship between explanatory variables and patent lifespan was insufficient. In this study, patent lifespan (number of renewals) is predicted based on the idea that patent lifespan represents the value of the patent. For the research, 4,033,414 patents applied between 1996 and 2017 and finally granted were collected from USPTO (US Patent and Trademark Office). To predict the patent lifespan, we use variables that can reflect the characteristics of the patent, the patent owner's characteristics, and the inventor's characteristics. We build four different models (Ridge Regression, Random Forest, Feed Forward Neural Network, Gradient Boosting Models) and perform hyperparameter tuning through 5-fold Cross Validation. Then, the performance of the generated models are evaluated, and the relative importance of predictors is also presented. In addition, based on the Gradient Boosting Model which have excellent performance, Accumulated Local Effects Plot is presented to visualize the relationship between predictors and patent lifespan. Finally, we apply Kernal SHAP (SHapley Additive exPlanations) to present the evaluation reason of individual patents, and discuss applicability to the patent evaluation system. This study has academic significance in that it cumulatively contributes to the existing patent life estimation research and supplements the limitations of existing patent life estimation studies based on linearity. It is academically meaningful that this study contributes cumulatively to the existing studies which estimate patent lifespan, and that it supplements the limitations of linear models. Also, it is practically meaningful to suggest a method for deriving the evaluation basis for individual patent value and examine the applicability to patent evaluation systems.

• Journal of the Society of Naval Architects of Korea
• /
• v.29 no.4
• /
• pp.114-131
• /
• 1992

In the present work, a two-dimensional boundary-value problem for a large amplitude motion is treated as an initial-value problem by satisfying the exact body-boundary and nonlinear free-surface boundary conditions. The present nonlinear numerical scheme is similar to that described by Vinje and Brevig(1981) who utilized the Cauchy's theorem and assumed the periodicity in the horizontal coordinate. In the present thesis, however, the periodicity in the horizontal coordinate is not assumed. Thus the present method can treat more realistic problems, which allow radiating waves to infinities. In the present method of solution, the original infinite fluid domain, is divided into two subdomains ; ie the inner and outer subdomains which are a local nonlinear subdomain and the truncated infinite linear subdomain, respectively. By imposing an appropriate matching condition, the computation is carried out only in the inner domain which includes the body. Here we adopt the nonlinear scheme of Vinje & Brevig only in the inner domain and respresent the solution in the truncated infinite subdomains by distributing the time-dependent Green function on the matching boundaries. The matching condition is that the velocity potential and stream function are required to be continuous across the matching boundary. In the computations we used, if necessary, a regriding algorithm on the free surface which could give converged stable solutions successfully even for the breaking waves. In harmonic oscillation problem, each harmonic component and time-mean force are obtained by the Fourier transform of the computed forces in the time domain. The numerical calculations are made for the following problems. $\cdot$ Forced harmonic large-amplitude oscillation(${\omega}{\neq}0,\;U=0$) $\cdot$ Translation with a uniform speed(${\omega}=0,\;U{\neq}0$) The computed results are compared with available experimental data and other analytical results.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.10a
• /
• pp.733-735
• /
• 2014

Example based Super Resolution(SR) is using the correspondence between the low and high resolution image from a database. This method uses only one image to estimate a high resolution image and can get the larger image than 2 times. Example based SR is proposed to solve the problem of classical SR. Neighbor embedding(NE) has been inspired by manifold learning method, particularly locally linear embedding. However, the poor generalization of NE decreases the performance of such algorithm. The sizes of local training sets are always too small to improve the performance of NE. We propose the advanced NE baesd on SVR having an excellent generalization ability to solve this problem. Given a low resolution image, we estimate a pixel in its high resolution version by using SVR based NE. Through experimental results, we quantitatively and qualitatively confirm the improved results of the proposed algorithm when comparing with conventional interpolation methods and NE.

• Journal of agricultural medicine and community health
• /
• v.30 no.2
• /
• pp.127-135
• /
• 2005

Objectives: This study was conducted to investigate the trend of tuberculosis mortality rate by years and by areas. Methods: We calculated raw and age-adjusted mortality rate of tuberculosis from 1995 to 2002. The calculation was based on the data from resident registration data and death certification registration data gathered by 232 basic local authority. We used direct age standardization method for calculating age-adjusted mortality rate. We compared patterns of change in tuberculosis mortality rate of metropolitan areas, cities, and countryside by determinating the comparability of medels to explore linear relationship. We also analyzed the data of mortality rate between urban and rural area by comparing ANOVA and post-hoc by two periods: one from 1995 to 1998, and the other from 1999 to 2002. Results: In national mortality rate, both raw and age-adjusted mortality rate showed negative linear relationship. However, the graph become more horizontal: the slope line is close to zero. From 1995 to 1998, countryside showed significantly higher age-adjusted mortality rate than in metropolitan areas and cities. Ever after considering more horizontal graph in national mortality rate, the data shows that the countryside still have significantly higher mortality rate from 1999 to 2002. In model diagnostic checking, metropolitan areas and cities showed apparently linear pattern on the decrease of age-adjusted mortality rate. Pattern of mortality rate in countryside was decreased initially, but became flat. Conclusions: Further research is necessary to explore the characteristics of quality of tuberculosis control program in rural area. Different approach and strategies should be considered to decrease tuberculosis mortality rate in rural areas.

• Korean Journal of Environment and Ecology
• /
• v.30 no.6
• /
• pp.1022-1031
• /
• 2016

Relationships of standing biomass with biodiversity and structural diversity were examined in the Quercus mongolica-dominated forest in Mt. Jeombong, Gangwon-do. We examined the standing biomass of the Q. mongolia community ($311.1ton{\cdot}ha^{-1}$) from 2004 to 2013, and the observed major species were Q. mongoilca, Carpinus cordata, Tilia amurensis whose standing biomasses were $206.3ton{\cdot}ha^{-1}$ (66.3%), $36.9ton{\cdot}ha^{-1}$ (11.9%), and $30.6ton{\cdot}ha^{-1}$ (9.8%), respectively. Although the number of Q. mongolica individuals was very small compared with total density, the reason that Q. mongolica showed the most biomass than other species is due to greater average diameter at breast height (DBH) and the higher number of $DBH{\geq}50cm$ individuals. We calculated the range of Shannon index (H') and Shannon evenness (J') in the Q. mongolica community, and they were gradually increased in time, showing 2.015~2.166, 0.673~0.736, respectively. Their H' and J' showed positive linear relationships with their standing biomass. This indicates that the spatial distribution of the standing biomass in Q. mongoilca community becomes more homogeneous with time and this homogenization appears in various species in the community. In addition, we estimated biomass-species index (BS) and abundance-biomass-speciesdiversity (ABS) and they also showed gradual increase in time, ranging from 3.746 to 3.811 and from 4.781 to 5.028, respectively. Their indices showed positive linear relationships with the standing biomass. This can be explained from the observations of variations in standing biomass with tree diameters as the differences in the average standing biomass in the community have reduced gradually in time. Moreover, it is expected that increase in the structure diversity of the Q. mongoilca community enhances the efficiency in carbon sequestration and productivity, so the community can be developed to a more sustainable ecosystem with more abundant resources. Thus, applications of uneven-aged plantations with considerations of local ecological properties can be a very efficient reforestation method to ensure stable support of biodiversity and productivity.

• Korean Journal of Community Nutrition
• /
• v.20 no.3
• /
• pp.220-235
• /
• 2015

Objectives: The purpose of this study was to investigate how clinical nutrition services is provided at a long term care hospital in Korea and to investigate job satisfaction levels of the clinical dietitians. Methods: Survey questionnaire was sent to dietitians working at a long term care hospital in Korea. The participating hospitals (n=240) were randomly selected from 1,180 long- term care hospitals using a stratified sampling method. A total of 134 long term care hospital s and 223 dietitians completed the survey of clinical nutrition service s and job satisfaction questionnaires The job satisfaction questionnaire included 27 job satisfaction questions on task, stability vision, working conditions, and relationship areas. Results: The average nutritional screening rate was 17.9% and the rate of computerized nutritional screening system was 9.7% in the participating hospitals. Nutritional intervention rate was only 3.2% of all patients. KOIHA (Korea Institute for Healthcare Accreditation) accreditated hospitals showed only 50% performance rate of nutrition service evaluation area. This shows that after achieving KOIHA accredition, many hospitals do not emphasize the performance of nutritional services. The job satisfaction scores in all four areas ranged from 2/5 to 3/5, implying generally low job satisfaction level in hospital dietitians. Linear regression analysis results showed that the "hospital adequacy grade" type was a significant predictor of job satisfaction level for two areas (working conditions & relationship). Conclusions: There is a need to provide proper standardized clinical nutrition services as a primary treatment and we observed large variations in the quality of nutritional service s in long term care hospitals. Therefore, government and local hospitals have to work on implementing nutritional programs and policies for improved service and care.

• Journal of Korea Water Resources Association
• /
• v.43 no.12
• /
• pp.1083-1091
• /
• 2010

Distributed hydrologic models typically require spatial estimates of precipitation interpolated from sparsely located observational points to the specific grid points. However, widely used estimation schemes fail to describe the realistic variability of daily precipitation field. We compare and contrast the performance of statistical methods for the spatial estimation of precipitation in two hydrologically different basins, and propose a two-step process for effective daily precipitation estimation. The methods assessed are: (1) Inverse Distance Weighted Average (IDW); (2) Multiple Linear Regression (MLR); (3) Climatological MLR; and (4) Locally Weighted Polynomial Regression (LWP). In the suggested simple two-step estimation process, precipitation occurrence is first generated via a logistic regression model before applying IDW scheme (one of the local scheme) to estimate the amount of precipitation separately on wet days. As the results, the suggested method shows the better performance of daily rainfall interpolation which has spatial differences compared with conventional methods. And this technique can be used for streamflow forecasting and downscaling of atmospheric circulation model effectively.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.6 no.3
• /
• pp.213-225
• /
• 2004

The number of tunnels are in fact increasing as a part of linear improvement project of general national highway and road enlargement and pavement project. Recently, collapses of portal slope are also occurring considerably, due to local raining from severe rain storm and abnormal weather. Accordingly, it was risen a necessity to efficiently respond to tunnel portal slope damage and maintenance in Korea and oversea nations. This paper is a basic proposal to execute a survey on the current status and state of the tunnel portal slopes that were already installed and are now being operated along general national highways, and also to execute state evaluation for the purpose of managing those effectively. As a research method, domestic tunnels were analyzed in accordance with geometrical shape such as access type, portal form, and tunnel type, etc. via field survey to analyze the types of tunnel portal slopes along national highways. State evaluation classification sheet is presented to divide classes for the danger state of the surveyed portal slopes, and then the related grades are divided. It is mainly aimed at classifying the tunnel portal slope along national highways with using this state evaluation, to use it as basic data so that continuous maintenance can be executed in the future in accordance with danger classes.