• 대한수학회보
• /
• 제58권3호
• /
• pp.537-557
• /
• 2021

In this paper, we investigate the curvature behavior of complete noncompact gradient expanding Ricci solitons with nonnegative Ricci curvature. For such a soliton in dimension four, it is shown that the Riemann curvature tensor and its covariant derivatives are bounded. Moreover, the Ricci curvature is controlled by the scalar curvature. In higher dimensions, we prove that the Riemann curvature tensor grows at most polynomially in the distance function.

• 대한수학회지
• /
• 제60권5호
• /
• pp.1023-1041
• /
• 2023

In this paper, motivated by the work of Q. S. Zhang in [25], we derive optimal Li-Yau gradient bounds for positive solutions of the f-heat equation on closed manifolds with Bakry-Émery Ricci curvature bounded below.

• 대한의용생체공학회:의공학회지
• /
• 제20권1호
• /
• pp.53-58
• /
• 1999

자기공명영상시스템에서 경사자계코일에 전류 펄스를 인가할 때 코일 주변의 도전성 구조물에 유도되는 와전류는 경사자계의 선형성을 열화 시킬 뿐만 아니라 경사자계의 파형을 왜곡시킨다. 특히 경사자계 파형의 왜곡 정도가 공간 위치에 따라 다를 경우 고속촬영법과 같은 고성능의 자기공명영상법을 실현하는 것이 어려워 진다. 본 연구에서는 촬영 공간 내 임의의 지점에서 와전류에 의해 경사자계 파형이 변형되는 정도를 알 수 있는 방법을 유한요소법을 이용해 구하였다. 경사자계코일에 정현파를 인가했을 때 임의의 지점에서 자계의 진폭 및 위상 특성이 주파수에 따라 어떻게 변하는지를 나타내는 와전류 영향 전달함수를 구하였고, 이 전달함수를 이용해 임의의 입력 전류 파형에 대한 출력 경사자계 파형을 예측하였다. 제안한 방법의 타당성을 검증하기 위해 실제로 측정한 경사자계 파형을 예측한 파형과 비교하여 제시하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권11호
• /
• pp.4355-4371
• /
• 2020

Phase retrieval, recovering a signal from phaseless measurements, is generally considered to be an NP-hard problem. This paper adopts an amplitude-based nonconvex optimization cost function to develop a new stochastic gradient algorithm, named stochastic reweighted phase retrieval (SRPR). SRPR is a stochastic gradient iteration algorithm, which runs in two stages: First, we use a truncated sample stochastic variance reduction algorithm to initialize the objective function. The second stage is the gradient refinement stage, which uses continuous updating of the amplitude-based stochastic weighted gradient algorithm to improve the initial estimate. Because of the stochastic method, each iteration of the two stages of SRPR involves only one equation. Therefore, SRPR is simple, scalable, and fast. Compared with the state-of-the-art phase retrieval algorithm, simulation results show that SRPR has a faster convergence speed and fewer magnitude-only measurements required to reconstruct the signal, under the real- or complex- cases.

• 수산해양교육연구
• /
• 제20권2호
• /
• pp.156-167
• /
• 2008

Meorological data have been collected to monitor the wetland area in Goheung bay since 2003 and four intensive observations were conducted to study effects of the atmospheric turbulence on the energy budget and the ecological changes. We improved an algorithm to estimate the sensible heat flux with routine data. The sensible heat flux estimated by gradient method was in good agreement with that measured by precision instruments such as surface layer scintillometer and ultrasonic anemometer. Diurnal variations of sensible heat flux showed analogous tendency to those of temperature gradient. When the vertical wind shear of horizontal wind components was weak, even though temperature gradient was strong, the gradient method underestimated the sensible heat flux. A compensation for the cloud will make this gradient method be a helpful tool to monitor the ecosystem without expensive instruments except for weak wind shear and temperature gradient.

• 한국안광학회지
• /
• 제9권2호
• /
• pp.223-231
• /
• 2004

기존에 이루어진 대부분의 AC/A비에 관한 연구들은 주로 Gradient AC/A비를 측정함으로써 진행되어 왔다. 본 연구는 Gradient AC/A비와 Calculated AC/A비를 함께 측정해 봄으로써 Gradient AC/A비 만을 측정했을 경우에는 나타나지 않는 근접성 폭주비 (Proximal Convergence/Accommodation)에 대한 고찰을 통해서 AC/A비의 임상 활용도를 높이는 데 그 목적이 있다. 124명의 검사대상은 최대시력을 단안 각각 1.0이상으로 원용교정을 하였으며 동적 검영법을 실시하여 조절기능의 이상 유무를 확인한 뒤 각각 Calculated AC/A비와 Gradient AC/A비를 측정하여 PC/A비를 구하였다. 연구결과 양안시 이상 그룹별 평균 Calculated AC/A비와 평균 Gradient AC/A비의 차이는 근접성 폭주 때문임을 보여주었다. 따라서, AC/A비의 임상 활용도를 높이기 위해서는 PC/A비를 포함한 근접성 (proximity)이라는 시기능변수 전반에 관한 지속적인 연구가 필요함을 확인할 수 있었다.

• 한국컴퓨터정보학회논문지
• /
• 제16권2호
• /
• pp.93-99
• /
• 2011

본 논문에서는 강화학습(RL : Reinforcement Learning) 환경 하에서 정책 기울기 값 기법을 사용하는 적응적인 QoS 라우팅 기법을 제안하였다. 이 기법은 기존의 강화학습 환경 하에 제공하는 기법에 비해 기대 보상값의 기울기 값을 정책에 반영함으로써 빠른 네트워크 환경을 학습함으로써 보다 우수한 라우팅 성공률을 제공할 수 있는 기법이다. 이를 검증하기 위해 기존의 기법들과 비교 검증함으로써 그 우수성을 확인하였다.

• 로봇학회논문지
• /
• 제14권4호
• /
• pp.258-263
• /
• 2019

Some animals have special sensing functions in order to find foods, home and mates. Instead of passively sensing, they discharge signals and then extract necessary information from the response. More importantly, they utilize the gradients of the sensed signal in order to find the destination or objects. In this paper this special strategy is formulated mathematically, i.e., the perturbation and the correlation based gradient estimation is developed. A stereo sensor system using temperature sensors mounted on motors is developed for verification. The proposed method can estimate the gradient of the measured value accurately. Using this method, the direction in the maximum measured value can be estimated accurately, and the position of the heat source can be estimated from the intersection of the directions estimated from both sensors.

• 대한의용생체공학회:학술대회논문집
• /
• 대한의용생체공학회 1998년도 추계학술대회
• /
• pp.92-93
• /
• 1998

In this study, we have established the technique to estimate the gradient waveforms distorted by the eddy current in MRI. After obtaining the eddy current effect transfer function using the finite element method, we have used the transfer function to estimate the output gradient waveforms at any points inside the imaging region. We also present experimental results to be compared with estimated ones.

• ETRI Journal
• /
• 제28권1호
• /
• pp.59-66
• /
• 2006

Learning behaviors of a radial basis function network (RBFN) using a singular value decomposition (SVD) and stochastic gradient (SG) algorithm, together named RBF-SVD-SG, for odor sensing systems are analyzed, and a fast training method is proposed. RBF input data is from a conducting polymer sensor array. It is revealed in this paper that the SG algorithm for the fine-tuning of centers and widths still shows ill-behaving learning results when a sufficiently small convergence coefficient is not used. Since the tuning of centers in RBFN plays a dominant role in the performance of RBFN odor sensing systems, our analysis is focused on the center-gradient variance of the RBFN-SVD-SG algorithm. We found analytically that the steadystate weight fluctuation and large values of a convergence coefficient can lead to an increase in variance of the center-gradient estimate. Based on this analysis, we propose to use the least mean square algorithm instead of SVD in adjusting the weight for stable steady-state weight behavior. Experimental results of the proposed algorithm have shown faster learning speed and better classification performance.