http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
전정 유모세포 통합 모델을 이용한 반강성 기전 기반 섬모번들 특성 추정에 관한 연구
김동영,홍기환,김규성,이상민,Kim, Dongyoung,Hong, Kihwan,Kim, Kyu-Sung,Lee, Sangmin 대한의용생체공학회 2013 의공학회지 Vol.34 No.4
In this paper hair bundle feature model and integration method for hair cell models were proposed. The proposed hair bundle feature model was based on spring-damper-mass model. Input of integrated vestibular hair cell model was frequency and output was interspike interval of hair cell that was reflected the feature of hair bundles. Irregular afferents that had a great gain variation showed reduction of negative stiffness section. Regular afferents that had a small gain variation, however, showed same feature with base negative stiffness feature. As a result, integrated vestibular hair cell model showed almost the same modeling data with experimental data in the modeled eleven frequency bands. It is verified that the proposed model is a good model for hair bundle feature modeling.
김동영 ( Dongyoung Kim ),이충희 ( Chung-hee Lee ) 한국정보처리학회 2014 한국정보처리학회 학술대회논문집 Vol.21 No.2
임의 숫자는 여러 분야에서 다양하게 사용되고 있으며,크게 True Random Number와 Pseudo Random Number로 구분지어 지는데,대부분의 경우 Pseudo Random Number를 사용하고 있다. 이 경우,동일한 Seed에 대해서는 항상 동일한 값을 반환하기 때문에,진정한 임의 숫자라고 하기는 어렵다. 본 논문에 서는 임의 숫자에 대한 기본 정의와 더불어 정지 영상을 이용하여 임의 숫자를 생성하는 방법에 대해 알아보고,기존의 Pseudo Random Number와의 차이점을 설명하도록 하겠다.
Integral Channel Feature를 이용한 보행자 검출 구현
김동영 ( Dongyoung Kim ),이충희 ( Chung-hee Lee ) 한국정보처리학회 2015 한국정보처리학회 학술대회논문집 Vol.22 No.1
최근 여러 매체에서 화두가 되고 있는 자율 주행 자동차나 Advanced driver assistance systems (ADAS)과 같은 분야에서 보행자 검출 기술은 핵심 요소 기술 중에 하나로 손꼽히고 있다. 특히, 인간의 인지 부하(Cognitive Load)를 고려했을 때, 주행 중에 발생할 수 있는 모든 사건을 다룬다는 것은 매우 어렵기 때문에, 앞서 언급한 방법의 도움을 받아 도로 주행 중에 발생 될 수 있는 인명 사고율을 줄이고자 하는데 그 목적이 있다. 본 논문에서는 Integral Channel Feature 를 사용하여 AdaBoost 알고리즘으로 보행자 검출을 위한 분류기를 구현하였다. 그 결과, INRIA 에서 제공되는 Pedestrian dataset 에서 Detection rate 는 97%이상, False positive 는 W에 정도로 나타났다.
Terrorism and Population Characteristic
Kim, Dongyoung(김동영),Kim, Young-Il(김영일) 한국인구학회 2017 한국인구학 Vol.40 No.1
본 논문은 European Social Survey (ESS)을 사용하여 샤를리 엡도 총격 테러 사건이 인구학적 특성 중 하나인 신뢰에 미치는 영향을 탐구한다. 인터뷰 날짜를 통해 처치 그룹과 통제 그룹을 구분하고, 이중 차분 모형을 주된 분석 방법으로 차용하여 저자들은 해당 총격 사건이 프랑스 거주민들의 신뢰 수준에 어떤 충격을 주었는지 분석한다. 프랑스 거주민들은 해당 사건으로 인해 신뢰 수준이 더 높아진 것으로 나타난다: 이는 일반 시민들에 대한 신뢰뿐만 아니라 국회와 경찰에 대한 신뢰까지 포함하며 회귀 단절 모형 분석에도 강건하다. 총격사건의 충격은 인구학적 특성인 이민자 여부, 동거 여부 그리고 소득 수준에 따라 다른 것으로 분석된다. We study the impact of Charlie Hebdo shooting on trust of French residents using the European Social Survey (ESS). This paper uses the difference-in-differences method to identify the impact of the terrorist attack on trust attitudes, exploiting the survey dates to differentiate the treatment group from the control group. French residents are more likely to have more favorable attitudes towards three different measures of trust: trust for people; parliament and police. The results are qualitatively robust to regression discontinuity design. We find that subgroups by immigration, cohabiting and income respond differently to the shooting.
A Critical Literature Review on the Data-Driven Language Learning Approach
Dongyoung Kim(김동영) 언어과학회 2019 언어과학연구 Vol.0 No.90
This paper examines the pedagogical advantages and disadvantages inherent to data-driven learning (DDL) and identifies principal characteristics associated with this approach. Most notably, DDL provides learners with authentic instances of the target language, and allows them to experience even specialized linguistic occurrences. DDL can not only provide educators with opportunities to implement communicative approaches to language instruction, it can also facilitate corrective function among students, thereby exemplifying a learner-centered approach. Nevertheless, there are challenges and limitations for DDL, which thus far hindered DDL’s widespread adoption. There have been concerns regarding the identification of authoritative sources of authentic language use. Besides, DDL’s nature may result in learner errors unless adequate time is dedicated to training and acclimatization. Firm empirical evidence regarding the efficiency of DDL is required if the technique is to achieve widespread acceptance.
시 · 공간 데이터를 활용한 머신러닝 기반 범죄예측모형 비교
김동영(Kim, Dongyoung),정성원(Jung, Sungwon) 대한건축학회 2021 대한건축학회논문집 Vol.37 No.1
With the advancement in computer performance and data analysis techniques, research using big data and machine learning is actively underway in various fields. However, regarding the domestic crime prediction research using machine learning, the current related studies are insufficient because disclosure of crime data is restricted and most of these studies predicted crimes by using a wide range of analysis units or by focusing on a few variables. To effectively distribute police power through practical crime prediction, it is necessary to predict the time and place of the crime. Therefore, in this study, we train machine learning models with 9,413 instances of actual theft crime data having temporal-spatial elements such as crime time and date, buildings, land-use, and CCTV. Thereby, we intend to provide a basis for future research and assist crime prevention activities practically by comparing the results of the prediction models. In this study, we divided the target land into 100 m square grids by using GIS and then inserted crime and temporal-spatial related variables. Subsequently, we trained the typical machine learning models such as random forest, decision trees, SVC, and K-NN, conducted crime prediction, and compared the results of the models. In the case of crime data, generally, an excessive amount of unbalanced data is present for the places where crimes did not occur compared to places where crimes occurred. Unbalanced data can result in noise and cause inaccurate predictions-these issues must be addressed. Therefore, in this study, we proposed a resampling method as an alternative to solve data imbalances and provide crime prediction with improved accuracy. The results of the comparison of the prediction performance of the models showed that the F1 score of the random forest model using the SMOTE method was high. This could be because the data loss of the SMOTE method is less than that of the under-sampling method and the random forest as an ensemble type model has an advantage in predicting data with various variables. We compared the influence of each variable by employing the feature importance function. Overall, the temporal-related variables showed high influence-among them, crimes occurred within one month showed the highest influence. Among the physical environment-related variables, first neighborhood living facility, retail store, and detached house were found to have high influence.