• STI Policy Review
• /
• 제7권2호
• /
• pp.16-34
• /
• 2016

This study aims to identify the chronological trajectory of university science parks (USPs) in China and to discuss the roles of government-driven science and technology (S&T) policies in the development of USPs and the future directions of these entities. Our study shows that USPs in China have undergone two development waves: The first from the late 1980s to the late 1990s, when research universities expected to directly participate in economic activities, and the second from 2000 when the Ministry of Science and Technology (MOST) and the Ministry of Education (MOE) jointly enacted the Proposed Regulation of State-level USPs Management to guide and regulate the development of USPs. The development trajectory highlights that USPs are effective platforms that link scientific research, knowledge spillovers and industrial system. However, Chinese USPs still need to confront some conundrums which may influence the processes and outcomes of UILs. Finally, we also summarize the major issues inherent in the development of USPs to guide policymakers to enact more effective policies.

• PNF and Movement
• /
• 제14권2호
• /
• pp.75-81
• /
• 2016

Purpose: This study used a comparative analysis to propose the effectiveness of using scapular and pelvic patterns of proprioceptive neuromuscular facilitation simultaneously (USPS) on the components of gait in patients with stroke. Methods: Subjects who signed a written consent form to participate in the experiment and attended voluntarily were divided randomly into two groups: the USPS group (12) and the control group (9). The USPS group received pattern training for 30 minutes a day, 3 days per week for 6 weeks. The control group received general physical exercise for the same duration. Gait ability was tested and analyzed before and after the 6 weeks of training with the GAITRite system. Results: The components of gait increased significantly in the USPS group after training (p < 0.05) and did not increase significantly in the control group. Conclusion: According to the results of this study, USPS training improves the components of gait more than general physical exercise. The USPS training method is effective for improving motor functions in patients with stroke, especially for gait rehabilitation.

• 전자통신동향분석
• /
• 제36권3호
• /
• pp.145-160
• /
• 2021

We provide trends on postal vehicles owned and operated by 10 overseas postal agencies, USPS, Postes Canada, Royal Mail, La Poste, Deutsche Post DHL, PostNL, Australia Post, China Post, Japan Post, and Thailand Post. This will be a useful asset to look into how the posts worldwide operate and manages their postal vehicles and service strategies. In this study, we provide the simple specifications of vehicles operated by each post, how to replace vehicles, how to service maintenance and repair, and modes of pick-up & delivery to be introduced soon. We expect this study to present appropriate policy implications for both postal agencies and logistics companies that are facing a major turning point in logistics to operate and evolve the delivery vehicles.

• 산업공학
• /
• 제18권1호
• /
• pp.104-116
• /
• 2005

In this paperm we provide a plan of construction, employment and management of the Korean postal address database and discuss the design of to-be models for it. For the purpose, we adopt and modify the information strategy planning methodology to fit it into our study. We examine the managerial environments and limitations of the postal services and define the requirements as a result. The U.S. Postal Service (USPS) was chosen as the best practice for a benchmarking and gap analysis with its advanced management of address information. After analyzing the current status of the postal business processes and information systems used, we classify the products and the services of USPS by the process to which they are applied and propose a deployment plan. In order to design a to-be model, we define the operating and managing functions for the address database and specify the scenario according to the functions. By managing the address database and supporting the postal services, it could be possible to make the business competitive and satisfy the customer demands.

• 전자통신동향분석
• /
• 제17권5호통권77호
• /
• pp.201-211
• /
• 2002

미국의 우편산업은 연 9천억 달러 규모의 시장을 형성하고 있으며, 보편적인 공공서비스 제공 및 상거래의 원동력 역할이라는 사회적인 책임을 충실히 수행해 오고 있다. 그러나 최근의 경제 하향세 및 인터넷, 이동통신을 비롯한 다른 통신 채널과의 경쟁 추세는, 우편산업도 수익성 및 고객 만족이라는 기업적인 목표를 추구해야 하며 산업의 경쟁력을 확보할 것을 요구하고 있다. 본 논문은 미국 우정공사(USPS)를 비롯한 우편산업계의 환경변화와 이에 대처하기 위한 구체적인 전략 및 활동 등의 최근 동향에 대해 정리한다.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2001년도 봄 학술발표논문집 Vol.28 No.1 (B)
• /
• pp.586-588
• /
• 2001

최근에는 순환 신경망의 생성모델이 비교사 학습에 관련하여 활발히 연구되고 있다. 이러한 형태의 신경망은 형태 추출이나 인식에 효과적으로 사용될 수 있는 반면 반복 loop를 사용하므로 대단히 많은 계산이 필요하다. 본 논문에서는 Oh와 Seung에 의해 제안된 상향전파(Up-propagation) network이라는 순환 신경망을 FPGA를 이용해서 구현하였다. 단층 신경망은 9개의 상층 neuron과 256개의 하층 neuron으로 구성되 있으며 4만 게이트의 FPGA 하나로 효과적으로 구현할 수 있다. pipeline된 곱셈기로 게산 속도를 향상시켰고 sigmoid 전달 함수는 유한 정밀도의 2차 다항식으로 근사될 수 있다. 구현된 하드웨어는 hand-written 숫자 영상인 USPS data를 재생하는데 사용되었으며 좋은 결과를 얻었다.

• 산업경영시스템학회지
• /
• 제36권4호
• /
• pp.45-58
• /
• 2013

Korean local governments have been planning for u-City construction with great efforts. Such efforts are needed in order to reflect the urban characteristics of each city, to promote local economic growth, and to have a sustainable operation model. Thus it is critical at an earlier stage to set a strategic direction based on urban environment analysis, which will lead to a feasible implementation of u-City. This paper derives issues and suggests improvement methods based on a case study on the USPs (u-City Strategy Planning's) of ten local governments. It is expected that the improvement methods will greatly reduce risks involved in u-City construction and contribute to better provided local-specific u-services.

• 전자통신동향분석
• /
• 제36권4호
• /
• pp.118-134
• /
• 2021

In this study, we provide trends on alternative fuel vehicles promoted by postal agencies in 10 countries in North America, the EU, and Asia. It describes the specifications of most EV vehicles in operation, plans, and strategies to replace existing internal combustion engines with new vehicles in the future and provides the current status of alternative vehicle charging networks in each postal agency. This paper will help postal agencies, logistics companies, automobile companies, motorcycle companies, and even individuals who want to use vehicles with alternative fuels, such as electric vehicles, on strategies to establish and implement before introducing and operating the vehicles.

• Journal of information and communication convergence engineering
• /
• 제21권2호
• /
• pp.139-144
• /
• 2023

Engineers prefer deep neural networks (DNNs) for solving computer vision problems. However, DNNs pose two major problems. First, neural networks require large amounts of well-labeled data for training. Second, the covariate shift problem is common in computer vision problems. Domain adaptation has been proposed to mitigate this problem. Recent work on adversarial-learning-based unsupervised domain adaptation (UDA) has explained transferability and enabled the model to learn robust features. Despite this advantage, current methods do not guarantee the distinguishability of the latent space unless they consider class-aware information of the target domain. Furthermore, source and target examples alone cannot efficiently extract domain-invariant features from the encoded spaces. To alleviate the problems of existing UDA methods, we propose the mixup regularization in adversarial discriminative domain adaptation (ADDA) method. We validated the effectiveness and generality of the proposed method by performing experiments under three adaptation scenarios: MNIST to USPS, SVHN to MNIST, and MNIST to MNIST-M.

• 대한전자공학회논문지SP
• /
• 제42권5호
• /
• pp.123-136
• /
• 2005

본 논문에서는 신경회로망보다 우수한 성능을 보이는 학습 이론인 SVM을 기반으로, 인간의 인지 과학에서 많은 연구가 이루어지고 있는 선택적 주의집중을 응용한 중첩 패턴 인식 시스템을 제안한다. 제안된 선택적 주의집중 모델은 SVM의 입력단에 주의집중층을 추가하여 SVM의 입력을 직접 변화시키는 학습을 하며 선택적 필터의 기능을 수행한다. 주의집중의 핵심은 학습을 멈추는 적절한 시점을 찾는 것과 그 시점에서 결과를 판단하는 주의집중 척도를 정의하는 것이다. 지지벡터는 주변에 존재하는 패턴들을 대표하는 표본이므로 입력 패턴이 초기상태일 때 주의집중을 하고자 하는 클래스의 가장 가까운 지지벡터를 기준으로 그 지지벡터와의 거리가 최소가 되었을 때 주의집중을 멈추는 것이 적절하다. 일반적인 주의집중을 적용하면 주의집중 척도를 정의하기가 난해해지기 때문에 변형된 입력이 원래 입력의 범위를 넘지 않는다는 제약조건을 추가하여 사용할 수 있는 정보의 폭을 넓히고 새로운 척도를 정의하였다. 이때 사용한 정보는 변형된 입력과 원래 입력의 유클리드 거리, SVM의 출력, 초기상태에 가장 가까웠던 히든뉴런의 출력값이다. 인식 실험을 위해 USPS 숫자 데이터를 사용하여 45개의 조합으로 중첩시켰으며, 주의집중을 적용시켰을 때 단일 SVM보다 인식 성능이 월등히 우수함을 확인하였고, 또한 제한된 주의집중을 사용하였을 때 일반적 주의집중을 이용하는 것 보다 성능이 더 뛰어났음을 확인하였다.