• ETRI Journal
• /
• v.43 no.4
• /
• pp.694-701
• /
• 2021

Catastrophic events cause casualties, damage property, and lead to huge social impacts. To build common standards and facilitate international communications regarding disasters, the relevant authorities in social management rank them in subjectively imposed terms such as direct economic losses and loss of life. Terrorist attacks involving uncertain human factors, which are roughly graded based on the rule of property damage, are even more difficult to interpret and assess. In this paper, we collected 114 183 open-source records of terrorist attacks and used a machine learning method to grade them synthetically in an automatic and objective way. No subjective claims or personal preferences were involved in the grading, and each derived common factor contains the comprehensive and rich information of many variables. Our work presents a new automatic ranking approach and is suitable for a broad range of gradation problems. Furthermore, we can use this model to grade all such attacks globally and visualize them to provide new insights.

• Journal of Information Processing Systems
• /
• v.17 no.3
• /
• pp.441-452
• /
• 2021

Federated learning provides an efficient integrated model for distributed data, allowing the local training of different data. Meanwhile, the goal of multi-task learning is to simultaneously establish models for multiple related tasks, and to obtain the underlying main structure. However, traditional federated multi-task learning models not only have strict requirements for the data distribution, but also demand large amounts of calculation and have slow convergence, which hindered their promotion in many fields. In our work, we apply the rank constraint on weight vectors of the multi-task learning model to adaptively adjust the task's similarity learning, according to the distribution of federal node data. The proposed model has a general framework for solving optimal solutions, which can be used to deal with various data types. Experiments show that our model has achieved the best results in different dataset. Notably, our model can still obtain stable results in datasets with large distribution differences. In addition, compared with traditional federated multi-task learning models, our algorithm is able to converge on a local optimal solution within limited training iterations.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.199-200
• /
• 2021

인터넷과 컴퓨팅 기술의 발전, 모바일 기기와 센서들의 진화, 소셜 네트워크의 출현 등으로 정보량은 급속도로 늘어나고 있다. 따라서 방대한 정보 속에서 의미있는 지식을 추출하기 위한 시스템의 기반 연구가 활발히 시도되고 있다. 본 논문에서는 텍스트 랭크를 사용한 중심 문장 추출을 통한 서비스와 사용자 이미지에 대한 한국어 OCR, 맞춤법 검사와 문장 생성을 가능케 하는 통합 한국어 처리 서비스 사이트를 구현함으로써, 신문 기사를 읽는 다수의 경제성을 확보했고, 한국어 처리의 편의성을 제공한다.

• Communications for Statistical Applications and Methods
• /
• v.31 no.6
• /
• pp.613-625
• /
• 2024

A general non-inferiority (NI) clinical trial is typically conducted using parametric testing methods with large samples. However, patient recruitment challenges often hinder rare disease trials, leading to enrollment failures. In this study, we introduce current parametric and nonparametric NI trial testing methods and propose modifications to enhance the performance of the nonparametric approach. Through a comprehensive simulation study with various sample sizes, data distributions, and sample ratios, we compare empirical levels and statistical powers as criteria for evaluating performance. Our findings indicate that the modified nonparametric methods outperformed the existing methods, particularly under conditions of small sample sizes and non-normal distributions, offering valuable insights for improving the reliability and sensitivity of NI trials in the context of rare diseases.

• Journal of the Korean Society of Industry Convergence
• /
• v.27 no.5
• /
• pp.1253-1260
• /
• 2024

This study analyzed psychological burnout among firefighters in fire suppression units and identified factors that influence it. Fire suppression work involves high levels of stress and repeated exposure to traumatic events, making psychological burnout particularly severe in this field. This burnout negatively impacts job performance and organizational efficiency. This study conducted an online survey of firefighters across South Korea, with 120 respondents, and analyzed the level of psychological burnout based on factors such as age, years of service, and rank. The results showed that the mean score for overall psychological burnout was 2.8 out of 5, indicating a moderate level of burnout among firefighters. Furthermore, personnel in lower ranks (firefighters, senior firefighters, fire sergeants, and fire lieutenants) experienced higher levels of psychological burnout compared to those in higher ranks (fire captains, deputy fire chiefs, etc.). These findings suggest the need for rank-specific burnout management strategies. Overall, the results of this study contribute to alleviating psychological burnout among firefighters, enhancing organizational efficiency, and strengthening public safety.

• Journal of KIISE:Software and Applications
• /
• v.29 no.3
• /
• pp.145-155
• /
• 2002

In content-based image retrieval (CBIR), most conventional approaches assume a linear relationship between different features and require users themselves to assign the appropriate weights to each feature. However, the linear relationship assumed between the features is too restricted to accurately represent high-level concepts and the intricacies of human perception. In this paper, a neural network-based image retrieval (NNIR) model is proposed. It has been developed based on a human-computer interaction approach to CBIR using a radial basis function network (RBFN). By using the RBFN, this approach determines the nonlinear relationship between features and it allows the user to select an initial query image and search incrementally the target images via relevance feedback so that more accurate similarity comparison between images can be supported. The experiment was performed to calculate the level of recall and precision based on a database that contains 1,015 images and consists of 145 classes. The experimental results showed that the recall and level of the proposed approach were 93.45% and 80.61% respectively, which is superior than precision the existing approaches such as the linearly combining approach, the rank-based method, and the backpropagation algorithm-based method.

• Journal of the Korean Statistical Society
• /
• v.23 no.1
• /
• pp.135-149
• /
• 1994

We discussed a comparison of distribution-free two-sample procedures based on placements or ranks. Iterative asymptotic distribution of both two-sample procedures is studies and small sample Monte Carlo simulation results are presented. Also, we proposed the Hodges-Lehmann type location estimator based on linear placement statistics.

• Journal of the Korean Data and Information Science Society
• /
• v.18 no.2
• /
• pp.411-418
• /
• 2007

Sample size calculation is very important in clinical trials. In this paper, we propose sample size calculation method for non-inferiority trials using sample size calculation method suggested by Wang et al.(2003) based on Wilcoxon's rank sum test. Also, sample size comparison between parametric method and proposed method are presented.

• Proceedings of the Korean Reliability Society Conference
• /
• 2000.11a
• /
• pp.363-363
• /
• 2000

We derive an asymptotic relation between the scores for the censored and uncensored observations for the untied value case among uncensored observations. With this relation, we show that two types of the linear rank statistics which are based on any consistent estimates of the distribution function, are asymptotically equivalent. Also, we discuss another asymptotic equivalence considered by Cuzick (1985).

• Journal of the Korean Statistical Society
• /
• v.26 no.2
• /
• pp.211-222
• /
• 1997

We briefly review three Raleigh type location tests based on direction vectors, which have been shown to be efficient when the distribution is unknown, skewed, or heavy-tailed. Then we calculate their test breakdown points and discuss the robustness of Randles multivariate sign test for one-sample.