• Journal of Korea Multimedia Society
• /
• v.21 no.12
• /
• pp.1363-1369
• /
• 2018

Breast cancer is one of the most common cancers in women worldwide. In Korea, breast cancer is most common cancer in women followed by thyroid cancer. The purpose of this study is to evaluate the possibility of using deep - run model for segmentation of breast masses and to identify the best deep-run model for breast mass segmentation. In this study, data of patients with breast masses were collected at Asan Medical Center. We used 596 images of mammography and 596 images of gold standard. In the area of interest of the medical image, it was cut into a rectangular shape with a margin of about 10% up and down, and then converted into an 8-bit image by adjusting the window width and level. Also, the size of the image was resampled to $150{\times}150$. In Deconvolution net, the average accuracy is 91.78%. In U-net, the average accuracy is 90.09%. Deconvolution net showed slightly better performance than U-net in this study, so it is expected that deconvolution net will be better for breast mass segmentation. However, because of few cases, there are a few images that are not accurately segmented. Therefore, more research is needed with various training data.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.10
• /
• pp.3608-3626
• /
• 2021

The detection of bolts is an important task in high-speed train inspection systems, and it is frequently performed to ensure the safety of trains. The difficulty of the vision-based bolt inspection system lies in small sample defect detection, which makes the end-to-end network ineffective. In this paper, the problem is resolved in two stages, which includes the detection network and cascaded classification networks. For small bolt detection, all bolts including defective bolts and normal bolts are put together for conducting annotation training, a new loss function and a new boundingbox selection based on the smallest axis-aligned convex set are proposed. These allow YOLOv3 network to obtain the accurate position and bounding box of the various bolts. The average precision has been greatly improved on PASCAL VOC, MS COCO and actual data set. After that, the Siamese network is employed for estimating the status of the bolts. Using the convolutional Siamese network, we are able to get strong results on few-shot classification. Extensive experiments and comparisons on actual data set show that the system outperforms state-of-the-art algorithms in bolt inspection.

• Journal of Distribution Science
• /
• v.19 no.8
• /
• pp.47-56
• /
• 2021

Purpose: Maritime sector is fundamental to international trade; there is no doubt that seafarers have played an essential role in maritime shipping and distribution science industry. Thus, this study uses Grey models to predict the number of seafarers in Vietnam expecting to provide a range of future seafarers. Research design, data and methodology: Statistics data are adopted for numbers of seafarers by Vietnam Maritime Administration categorizing into three types: Officers at Management level, Officers at Operational level and Navigation - Engine officer cadet. Results: The results have showed that a lack of qualified seafarers in the distribution industry, which has become a global issue and Vietnam is facing challenges of providing enough supply of seafarers in the next few years. Since there has been a concern of the unbalance between demand and supply of seafarers, researches in maritime sector needs a high accuracy in forecasting the number of available qualified seafarers in Vietnam. Conclusion: This method can be applied to predict numbers of other human resources in transportation, distribution and/or logistics industries when the information is poor and insufficient. The next few years are predicted to witness a downtrend in sailors - oilers which leads to the fact that the total number of available seafarers is decreased.

• International Journal of Knowledge Content Development & Technology
• /
• v.14 no.1
• /
• pp.53-73
• /
• 2024

The purpose of the study is to investigate the development of institutional repositories in university libraries in Nigeria. The study adopted a survey research design. Online questionnaire, IRs investigation, and interview methods were used to collect data from the 21 university libraries that have developed IRs in Nigeria. The study revealed that only few universities have successfully developed open access IRs to preserve and manage their intellectual outputs emanating from their universities. Contents such as journal articles, theses/dissertations, and conference / workshop papers were found to be the most popular contents that are hosted in the IRs. The interview revealed that although few respondents mentioned having IR policy statements in areas such as access policy, submission policy, preservation policy, content policy and copyright policy, the majority mentioned not having any defined policy in their IR. Also in the interview, inadequate fund, challenge of collecting contents for the IR, shortage of skilled ICT personnel, and inadequate facilities were the most mentioned challenges encountered in the development of IRs in Nigeria. Findings from this study will inform University Librarians, university management and policy makers on the need to provide the necessary infrastructure and formulate policies for smooth development of institutional repositories to make research visible globally. The results will therefore provide important data and insight into the development of institutional repositories in university libraries in the context of developing countries.

• Nuclear Engineering and Technology
• /
• v.56 no.6
• /
• pp.2163-2173
• /
• 2024

This study introduces a transient analysis module developed for RAST-V and validates it using the Kalinin-3 benchmark problem. For the benchmark analysis, RAST-V standalone and STREAM/RAST-V calculations were performed. STREAM supplies the few-group constants and RAST-V conducts a 3D core simulation utilizing few-group cross-sectional data. To improve accuracy, the main solver was developed based on the advanced semi-analytic nodal method. To evaluate the computational capability of the transient analysis module in RAST-V, Kalinin-3 benchmark is employed. Kalinin-3 represents a coolant transient benchmark that offers experimental data during the deactivation of the Main Circulation Pumps. Consequently, the transient calculations reflected the changes in the reactor flow rate. Benchmark comprising steady-state and transient calculations. During the steady state, the STREAM/RAST-V combination demonstrated a 30 ppm root mean square difference from 0 to 128.50 EFPD. For the transient calculations, STREAM/RAST-V showed power differences within ±7 % over a range of 0-300 s. Axial offset differences were within ±3 %, and the RMS difference in radial power ranged within 2.596 % at both 0 and 300 s. Overall, this study effectively demonstrated the newly developed transient solver in RAST-V and validated it using the Kalinin-3 benchmark problem.

• Proceedings of the Korea Contents Association Conference
• /
• 2015.05a
• /
• pp.33-34
• /
• 2015

Continuous improvements of network infrastructures and mobile data offloading strategies are among the solutions of cellular providers to cope with the increase in mobile data demand. These options requires a lot of cost and time to implement. Few researches have been conducted to assess the applicability of Pocket Switched Network (PSN) to support mobile data offloading. In this paper, we present a PSN mobile data-offloading scheme that utilizes mobile users with available connectivity to deliver content-aware data to other mobile users. This paper also aims to evaluate the applicability and feasibility of PSN routing schemes to improve the current strategies in mobile data offloading. The simulation results show admirable results in terms of message delivery and latency.

• Journal of Korean Institute of Industrial Engineers
• /
• v.32 no.2
• /
• pp.133-140
• /
• 2006

Large signal type data sets are difficult to classify, especially if the data sets are non-stationary. In this paper, large signal type and non-stationary data sets are wavelet transformed so that distinct features of the data are extracted in wavelet domain rather than time domain. For the classification of the data, a few wavelet coefficients representing class properties are employed for statistical classification methods : Linear Discriminant Analysis, Quadratic Discriminant Analysis, Neural Network etc. The application of our wavelet-based feature selection method to a mass spectrometry data set for ovarian cancer diagnosis resulted in 100% classification accuracy.

• Communications for Statistical Applications and Methods
• /
• v.23 no.6
• /
• pp.555-562
• /
• 2016

Interval censored data often occur in an observational study where the subject is followed periodically. Instead of observing an exact failure time, two inspection times that include it are available. There are several methods to analyze interval censored failure time data (Sun, 2006). However, in the presence of competing risks, few methods have been suggested to estimate covariate effect on interval censored competing risk data. A sub-distribution hazard model is a commonly used regression model because it has one-to-one correspondence with a cumulative incidence function. Alternatively, Klein and Andersen (2005) proposed a pseudo-value approach that directly uses the cumulative incidence function. In this paper, we consider an extension of the pseudo-value approach into the interval censored data to estimate regression coefficients. The pseudo-values generated from the estimated cumulative incidence function then become response variables in a generalized estimating equation. Simulation studies show that the suggested method performs well in several situations and an HIV-AIDS cohort study is analyzed as a real data example.

• Communications for Statistical Applications and Methods
• /
• v.27 no.4
• /
• pp.397-412
• /
• 2020

In this paper, a notion of data depth is used to propose nonparametric multivariate two sample tests for difference between scale parameters. Data depth can be used to measure the centrality or outlying-ness of the multivariate data point relative to data cloud. A difference in the scale parameters indicates the difference in the depth values of a multivariate data point. By observing this fact on a depth vs depth plot (DD-plot), we propose nonparametric multivariate two sample tests for scale parameters of multivariate distributions. The p-values of these proposed tests are obtained by using Fisher's permutation approach. The power performance of these proposed tests has been reported for few symmetric and skewed multivariate distributions with the existing tests. Illustration with real-life data is also provided.

• Journal of the Korea Fashion and Costume Design Association
• /
• v.9 no.1
• /
• pp.123-140
• /
• 2007

Fashion magazine advertising is the most excellent source of information in predicting the fashion trend. It plays a pivotal role in setting a direction for the fashion trend in the upcoming season. The purpose of this study was to review by photos of the $pr{\hat{e}}t-{\acute{a}}-porter$ collection shown in LONDON, NEW YORK, MILAN, and PARIS during 2002-2006 A/W and 2003-2005 S/S seasons, being focused on such fashion items published as coat, dress, one-piece, two-pieces(jacket+skirt/pants, blouse+skirt/pants). In the results of this study, designers presented coat(n=144) chiefly, blouse+pants(n=29) were presented few during 2002-2003 A/W seasons. During 2003-2004 A/W seasons one-piece(n=156) was looking bullish, blouse+pants(n=34) were declining. Dress(n=149) was presented mostly, blouse+pants(n=17) was presented few during 2004-2005 A/W seasons. During 2005-2006 A/W seasons coat(n=180) was revived, blouse+pants(n=26) were presented lowly. Therefore designers presented coat(n=605, 28.4%) extremely much during 2002-2006 A/W seasons. Designers presented one-piece(n=109) much, jacket+pants(n=22) were presented few in 2003 S/S. In 2004 S/S seasons one-piece(n=167) was presented vastly different than jacket+pants(n=42). Also one-piece(n=152) was presented mostly, blouse+pants(n=48) was presented few in 2005 S/S seasons. During 2003-2005 S/S seasons one-piece(n=428, 28.2%) was presented most. And designers in these four world fashion centers didn't prefer blouse+pants in A/W seasons and jacket+pants in S/S seasons. Based on the above findings, it could be confirmed that the fashion items trend was almost similar among the four collections of $pr{\hat{e}}t-{\acute{a}}-porter$. The results of this study suggest that the fashion collections are the most reliable information sources for fashion product planning. Lastly, it is hoped that this study will provide for some useful basic data for domestic fashion businesses in producing fashion items.