• 방송공학회논문지
• /
• 제18권3호
• /
• pp.401-408
• /
• 2013

본 논문에서는 영상 내의 문맥 특징(context feature)과 외형 특징(appearance feature)을 함께 학습함으로써 의료영상 내의 비슷한 외형 특징을 가지는 장기들을 자동으로 검출하는 기법을 제안한다. 기존 검출 기법들은 외형 특징 정보만을 학습하여 분류기(classifier)를 생성하였기 때문에 의료영상 내에 외형이 비슷한 장기들이 다수 포함되어 있는 경우 검출 오류가 발생하였다. 제안하는 기법은 외형 특징을 이용하여 학습된 분류기를 통해 얻은 확률 값들을 바탕으로 관심 복셀(voxel) 주변의 확률 분포 특징을 반복적으로 학습함으로써 문맥 정보를 포함하는 분류기를 생성한다. 또한, 실험 단계(test stage)에서 '지역 기반 투표 방식'(region based voting scheme)을 도입함으로써 효율성과 정확성을 향상시킨다. 제안하는 기법의 성능 평가를 위해 SKI10 무릎 관절 데이터 셋 내에서 외형 특징이 비슷한 대퇴골(femur)과 경골(tibia)을 검출하는 실험을 진행하였다. 실험 결과를 통해 제안하는 기법이 외형 특징만을 이용했던 검출 기법에 비해 개선된 검출 성능을 보이고 있음을 확인할 수 있었다.

• International Journal of Quality Innovation
• /
• 제6권3호
• /
• pp.95-120
• /
• 2005

Utilization of information and communication technologies is commonly accepted as important to value creation in the knowledge economy. Nevertheless, empirical findings from our business case studies often show that while sophisticated technological tools may be developed, the potentials are not realized. It is evident that technology is subject to adaptive and emergent strategies of use, diverging from the original intention. Within this space of opportunities, we elaborate the importance of constructing strategic concepts as communication tools to support organisational implementation of technologies. We use the concept of organisational implementation as a way of taking the technology into use in order to support changes and value creation in the user organisation. In this paper we present our findings related to how use and experiences are conditioned by the users' expectations. We have conducted a business case study in order to understand and explore how users employ and use a particular wireless technology infrastructure. On behalf of the infrastructure vendor, we have studied three different organisations that use this technology. The overall research goal of our joint research project was to find out what is good use and for whom. We find that users struggle to go beyond the expectations they had when they were conceptualising and telling us about their practice. We have good indications that a narrowed consciousness was also conditioning the users' use of the technology. In this paper we draw the conclusion that technological implementations towards changing work practices and value creation must not be viewed by the company solely as a knowledge acquisition process, but as a process of knowledge creation. Organisational implementation is an ongoing process, a learning process at both the organisational and individual level. Flexible tools and technologies are constituted and shaped in interaction and communication in the workplace. Based on that knowledge, we build up an argument for an organisational implementation framework, including strategic discussions, learning spaces, and concept constructions.

• 한국안전학회지
• /
• 제24권3호
• /
• pp.71-78
• /
• 2009

During a learning process, a human being is assumed to experience knowledge-based behaviors, rule-based behaviors, and skill-based behaviors sequentially if Rasmussen was right. If any psycho-physiological symptom to those different levels can be obtained, it can be useful as a measure whether a human being is fully trained and has gotten a skill in his work. Therefore, this study aimed to draw relationships between human performance measures and psycho-physiological measures while committing a computer-simulated pointing task by utilizing the power spectrum technique of EEG data, especially with the ratio of relative beta-to-alpha band power. The result showed that, during correct responses, the ratio came to stabilize as all the performance data went stable. However, response time was not a simple linear function of task difficulty level only, but a joint function of task characteristics as well as behavior levels. Comparing relative band power ratios from errors and correct responses, activated states of one's brain could be explained, and characteristics of the task could understood. To tell that of pointing task, correlations around C3, C4, P3, P4 and 01, 02 area were significant and high in correct response cases whereas most correlation coefficients went down in error cases standing for imbalance of psycho-motor functions. Though task difficulty was the only one factor that could influence on relative band power ratio with statistical significance, it should be comprehended to mean a different way of expression indicating task characteristics since at least error-some situation could be explained with the help of relative band power ratio that absolute band power failed.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권10호
• /
• pp.3955-3971
• /
• 2020

Due to the significant increase in the use of autonomous car technology, it is essential to integrate this technology with high-precision digital map data containing more precise and accurate roadway information, as compared to existing conventional map resources, to ensure the safety of self-driving operations. While existing map technologies may assist vehicles in identifying their locations via Global Positioning System, it is however difficult to update the environmental changes of roadways in these maps. Roadway vision algorithms can be useful for building autonomous vehicles that can avoid accidents and detect real-time location changes. We incorporate a hybrid architectural design that combines unsupervised classification of vision data with supervised joint fusion classification to achieve a better noise-resistant algorithm. We identify, via a deep learning approach, an intelligent hybrid fusion algorithm for fusing multimodal vision feature data for roadway classifications and characterize its improvement in accuracy over unsupervised identifications using image processing and supervised vision classifiers. We analyzed over 93,000 vision frame data collected from a test vehicle in real roadways. The performance indicators of the proposed hybrid fusion algorithm are successfully evaluated for the generation of roadway digital maps for autonomous vehicles, with a recall of 0.94, precision of 0.96, and accuracy of 0.92.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2007년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.315-316
• /
• 2007

This paper presents a new methodology to improve movement database for a humanoid robot. The database is initially full of human motions so that the kinetics characteristics of human movement are immanent in it. then, the database is updated to the pseudo-optimal motions for the humanoid robot to perform more natural motions, which contain the kinetics characteristics of robot. for this, we use the evolutionary algorithm. the methodology consists of two processes : (1) the offline imitation learning of human movement and (2) the online generation of natural motion. The offline process improve the initial human motion database using the evolutionary algorithm and inverse dynamics-based optimization. The optimization procedure generate new motions using the movement primitive database, minimizing the joint torque. This learning process produces a new database that can endow the humanoid robot with natural motions, which requires minimal torques. In online process, using the linear combination of the motion primitive in this updated database, the humanoid robot can generate the natural motions in real time. The proposed framework gives a systematic methodology for a humanoid robot to learn natural motions from human motions considering dynamics of the robot. The experiment of catching a ball thrown by a man is performed to show the feasibility of the proposed framework.

• Journal of Information Processing Systems
• /
• 제8권4호
• /
• pp.575-588
• /
• 2012

Exploiting unlabeled text data with a relatively small labeled corpus has been an active and challenging research topic in text mining, due to the recent growth of the amount of biomedical literature. Biomedical named-entity recognition is an essential prerequisite task before effective text mining of biomedical literature can begin. This paper proposes an Active Co-Training (ACT) algorithm for biomedical named-entity recognition. ACT is a semi-supervised learning method in which two classifiers based on two different feature sets iteratively learn from informative examples that have been queried from the unlabeled data. We design a new classification problem to measure the informativeness of an example in unlabeled data. In this classification problem, the examples are classified based on a joint view of a feature set to be informative/non-informative to both classifiers. To form the training data for the classification problem, we adopt a query-by-committee method. Therefore, in the ACT, both classifiers are considered to be one committee, which is used on the labeled data to give the informativeness label to each example. The ACT method outperforms the traditional co-training algorithm in terms of f-measure as well as the number of training iterations performed to build a good classification model. The proposed method tends to efficiently exploit a large amount of unlabeled data by selecting a small number of examples having not only useful information but also a comprehensive pattern.

• International Journal of Computer Science & Network Security
• /
• 제22권4호
• /
• pp.193-202
• /
• 2022

The classification of hyperspectral imagery (HSI) is essential in the surface of earth observation. Due to the continuous large number of bands, HSI data provide rich information about the object of study; however, it suffers from the curse of dimensionality. Dimensionality reduction is an essential aspect of Machine learning classification. The algorithms based on feature extraction can overcome the data dimensionality issue, thereby allowing the classifiers to utilize comprehensive models to reduce computational costs. This paper assesses and compares two HSI classification techniques. The first is based on the Joint Spatial-Spectral Stacked Autoencoder (JSSSA) method, the second is based on a shallow Artificial Neural Network (SNN), and the third is used the SVM model. The performance of the JSSSA technique is better than the SNN classification technique based on the overall accuracy and Kappa coefficient values. We observed that the JSSSA based method surpasses the SNN technique with an overall accuracy of 96.13% and Kappa coefficient value of 0.95. SNN also achieved a good accuracy of 92.40% and a Kappa coefficient value of 0.90, and SVM achieved an accuracy of 82.87%. The current study suggests that both JSSSA and SNN based techniques prove to be efficient methods for hyperspectral classification of snow features. This work classified the labeled/ground-truth datasets of snow in multiple classes. The labeled/ground-truth data can be valuable for applying deep neural networks such as CNN, hybrid CNN, RNN for glaciology, and snow-related hazard applications.

• Genomics & Informatics
• /
• 제20권2호
• /
• pp.23.1-23.9
• /
• 2022

A survival prediction model has recently been developed to evaluate the prognosis of resected nonmetastatic pancreatic ductal adenocarcinoma based on a Cox model using two nationwide databases: Surveillance, Epidemiology and End Results (SEER) and Korea Tumor Registry System-Biliary Pancreas (KOTUS-BP). In this study, we applied two machine learning methods-random survival forests (RSF) and support vector machines (SVM)-for survival analysis and compared their prediction performance using the SEER and KOTUS-BP datasets. Three schemes were used for model development and evaluation. First, we utilized data from SEER for model development and used data from KOTUS-BP for external evaluation. Second, these two datasets were swapped by taking data from KOTUS-BP for model development and data from SEER for external evaluation. Finally, we mixed these two datasets half and half and utilized the mixed datasets for model development and validation. We used 9,624 patients from SEER and 3,281 patients from KOTUS-BP to construct a prediction model with seven covariates: age, sex, histologic differentiation, adjuvant treatment, resection margin status, and the American Joint Committee on Cancer 8th edition T-stage and N-stage. Comparing the three schemes, the performance of the Cox model, RSF, and SVM was better when using the mixed datasets than when using the unmixed datasets. When using the mixed datasets, the C-index, 1-year, 2-year, and 3-year time-dependent areas under the curve for the Cox model were 0.644, 0.698, 0.680, and 0.687, respectively. The Cox model performed slightly better than RSF and SVM.

• 한국과학교육학회지
• /
• 제41권4호
• /
• pp.297-310
• /
• 2021

이 연구에서는 교과교육전문성 함양을 목적으로 모인 과학교사 PLC를 학교안 PLC와 학교밖 PLC로 구분하여 그 특징을 탐색하였다. 이를 위하여 과학중점학교 7개교에 속한 과학교사 PLC와 학교밖 과학교사 PLC 3개 단체에 속한 과학교사들을 대상으로 학습공동체로서의 정체성과 동인, 활동 내용, 운영 성과, 부족하거나 지원이 필요한 부분 등에 대하여 초점집단 심층면담을 실시하였다. 주요 연구결과를 살펴보면, 학교안 과학과 PLC와 학교밖 과학과 PLC 모두 교수·학습과 관련된 전문성 개발을 위해 학습공동체를 구성한 것으로 나타났다. 연구에 참여한 학교안과 학교밖의 과학교사 학습공동체들은 모두 구성원들이 공통으로 경험하는 쟁점 문제들을 중심으로, 호혜적 참여를 통해 협력관계를 형성하는 '실천공동체'의 특징을 나타내었다. 한편, 쟁점 문제에서 학교안의 경우는 과학 교과 교육과정 재구성과 같은 미시적인 문제에, 학교밖의 경우는 교사로서의 전문성 향상과 같은 거시적인 문제에 천착하는 차이점을 나타내었다. 학습공동체로서의 활동과 역할의 경우 학교안과 밖의 과학교사 학습공동체들은 협업적 전문성 개발, 초임교사 멘토링과 같은 상호교학 등의 공통점을 나타내었다. 학습공동체의 영향력과 운영 성과의 경우, 학습공동체는 교사 자신뿐만 아니라, 학생과 학교문화에까지 긍정적인 영향을 미치는 것으로 나타났다. 학습공동체 지원과 관련하여 공동체 운영을 위한 물리적 여건 개선과 함께 PLC 운영을 위한 프로토콜 제공, 대학과의 연계를 통한 공동 연구나 재교육 등과 같은 소프트웨어 지원이 필요한 것으로 나타났다. 특히 대학과 PLC의 공동연구는 '탐구 공동체'로서 교사 학습공동체가 추구해야 할 미래지향적인 방향을 보여준다. 연구 결과를 토대로 과학과 PLC의 원활한 운영을 위한 적극적인 지원의 필요성, 과학과 PLC와 대학과의 협력 체제 구상의 필요성, 과학중점학교의 과학과 PLC의 운영 성과를 중·고등학교 과학교사 PLC에 확산하는 방안 등을 제안하였다.

• 근관절건강학회지
• /
• 제14권2호
• /
• pp.137-148
• /
• 2007

Purpose: This study was performed to evaluate of practicum by using standardized patient(SP) for nursing assessment. Method: This study had 2 steps. The 1st-step was pre-intervention stage including selection of a learning title, formation of case scenario, training of SP and developing the evaluation tools for students' clinical competence to assessment, 6 categories 29 items. The 2nd-step consisted of intervention and evaluation stages. 34 nursing students divided 2 groups participated in assessing the SP. Evaluation of each group was performed by 2 nursing professors. All students recorded their feelings after assessing the SP. The SP also evaluated about nursing students' attitude toward the SP. Results: ICC(Interclass correlation coefficient) between 2 groups was over 0.7 all categories. Students' assessing score(range 0-1) was muscular-joint function status(0.41), nutritional status(0.39), history taking(0.38), IADL(0.18), ADL(0.15), and emotional status (0.07). The mean scores of the nursing students' attitude by SP was 4.03(range 1-6). Also most students showed positive reactions to the education using SP because they had the chance to experience what they could not practice in clinical setting. Conclusion: The evaluation tool revealed high reliability. Nursing students' clinical competence was below average. But they took a good attitude to SP. We recommended further research using SP with various disease.