• 한국공간정보시스템학회:학술대회논문집
• /
• 한국공간정보시스템학회 2002년도 춘계학술대회 논문집
• /
• pp.45-52
• /
• 2002

Artificial neural networks (ANN) have been successfully used for classifying remotely sensed imagery. However, ANN still is not the preferable choice for classification over the conventional classification methodology such as the maximum likelihood classifier commonly used in the industry production environment. This can be attributed to the ANN characteristic built-in stochastic process that creates difficulties in dealing with unequally represented training classes, and its training performance speed. In this paper we examined some practical aspects of training classes when using a back propagation neural network model for remotely sensed imagery. During the classification process of remotely sensed imagery, representative training patterns for each class are collected by polygons or by using a region-growing methodology over the imagery. The number of collected training patterns for each class may vary from several pixels to thousands. This unequally populated training data may cause the significant problems some neural network empirical models such as back-propagation have experienced. We investigate the effects of training over- or under- represented training patterns in classes and propose the pattern repopulation algorithm, and an adaptive alpha adjustment (AAA) algorithm to handle unequally represented classes. We also show the performance improvement when input patterns are presented in random fashion during the back-propagation training.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2008년도 추계학술대회A
• /
• pp.894-899
• /
• 2008

Measured sensor datum from a quadruped robotics is commonly used for recognizing physical environment information which controls the posture of robotics. We can advance the ambulation with this sensed information and need to synthesize various sensors for obtaining accurate data, but most of these sensors are expensive and require excessive load for the operation. Those defects can be serious problem when it comes to the prototype's practicality and mass production, and maintenance of the system. This paper suggests virtual sensor technology for avoiding previous defects and presents ways to apply a theory to a walking robotics through virtual sensor information which is trained with several kinds of actual sensor information from the prototype system; the general algorithm is initially based on the neural network theory of back propagation. In specific, we verified a possibility of replacing the virtual sensor with the actual one through a reaction force measurement experiment.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2007년도 추계종합학술대회
• /
• pp.981-984
• /
• 2007

문자 인식이란 시각 정보를 통하여 문자를 인식하고 의미를 이해하는 것으로 인간의 능력을 컴퓨터로 실현하는 패턴인식의 한 분야이다. 본 논문에서는 문자 인식 중 가장 많이 사용되고 있는 숫자 검출과 인식을 소개하고자 한다. 또한 숫자 인식을 위해서 인간의 두뇌를 모델로 하여 만들어진 신경망에 대한 기본적인 원리와 신경망의 학습을 위한 역 전파(Back propagation) 알고리즘에 대하여 알아보고자 한다.

• International Journal of Naval Architecture and Ocean Engineering
• /
• 제4권4호
• /
• pp.374-385
• /
• 2012

The indoor noise of a ship is usually determined using the A-weighted sound pressure level. However, in order to better understand this phenomenon, evaluation parameters that more accurately reflect the human sense of hearing are required. To find the level of the satisfaction index of the noise inside a naval vessel such as "Loudness" and "Annoyance", psycho-acoustic evaluation of various sound recordings from the naval vessel was performed in a laboratory. The objective of this paper is to develop a single index of "Loudness" and "Annoyance" for noise inside a naval vessel according to a psycho-acoustic evaluation by using psychological responses such as Noise Rating (NR), Noise Criterion (NC), Room Criterion (RC), Preferred Speech Interference Level (PSIL) and loudness level. Additionally, in order to determine a single index of satisfaction for noise such as "Loudness" and "Annoyance", with respect to a human's sense of hearing, a back-propagation neural network is applied.

• Journal of information and communication convergence engineering
• /
• 제2권3호
• /
• pp.177-181
• /
• 2004

In this paper, we suggest an Iris recognition system with an excellent recognition rate and confidence as an alternative biometric recognition technique that solves the limit in an existing individual discrimination. For its implementation, we extracted coefficients feature values with the wavelet transformation mainly used in the signal processing, and we used neural network to see a recognition rate. However, Scale Conjugate Gradient of nonlinear optimum method mainly used in neural network is not suitable to solve the optimum problem for its slow velocity of convergence. So we intended to enhance the recognition rate by using Levenberg-Marquardt Back-propagation which supplements existing Scale Conjugate Gradient for an implementation of the iris recognition system. We improved convergence velocity, efficiency, and stability by changing properly the size according to both convergence rate of solution and variation rate of variable vector with the implementation of an applied algorithm.

• 한국콘텐츠학회:학술대회논문집
• /
• 한국콘텐츠학회 2014년도 춘계 종합학술대회 논문집
• /
• pp.31-32
• /
• 2014

• 대한자원환경지질학회:학술대회논문집
• /
• 대한자원환경지질학회 2001년도 춘계 공동학술발표회
• /
• pp.181-184
• /
• 2001

• 전자공학회논문지CI
• /
• 제44권2호
• /
• pp.11-18
• /
• 2007

신경회로망의 학습에 널리 사용되고 있는 오차역전파 알고리즘은 최급하강법을 기초로 하고 있기 때문에 초기값에 따라서는 극소값에 떨어지거나, 신경회로망을 학습시킬 때 중간층 유닛수를 얼마로 설정하는 등의 문제점이 있다. 따라서 이러한 문제점을 해결하기 위하여, 본 논문에서는 3비트 패리티 판별을 위하여 신경회로망의 학습에 교차법, 돌연변이법에 새로운 기법을 도입한 개량형 유전적 알고리즘을 제안한다. 본 논문에서는 세대차이, 중간층 유닛수의 차이, 집단의 개체수의 차이에 대하여 실험을 실시하여, 본 방식이 학습 속도의 면에서 유효하다는 것을 나타낸다.

• 소음진동
• /
• 제4권1호
• /
• pp.59-69
• /
• 1994

A new LMS algorithm titled constrained LMS' is proposed for problems with constrained structure. The conventional LMS algorithm can not be used because it destroys the constrained structures of the weights or parameters. Proposed method uses error-back propagation, which is popular in training neural networks, for error minimization. The illustrative examplesare shown to demonstrate the applicability of the proposed algorithm.

• 한국산업정보학회논문지
• /
• 제5권3호
• /
• pp.117-124
• /
• 2000

시스톨릭 어레이는 간단한 연산능력을 가진 처리요소들이 정규적이고 국부적인 통신 선들로 연결된 병렬처리 시스템이다. 시스톨릭 어레이는 인공신경망에서 고밀하게 연결된 뉴런으로 인하여 발생하는 뉴런간의 복잡한 통신 문제를 해결하는 가장 좋은 방법 중의 하나로 알려져 있다. 본 논문에서는 주어진 뉴런수에 적합한 역전파 인공신경망을 자동으로 생성하는 시스톨릭 어레이 시뮬레이터를 설계하고 구현한다. 시뮬레이터의 애니메이션 기법을 이용하여, 설계된 시스틀릭 어레이 상에서의 역전파 알고리즘의 실행 상황을 사용자들이 단계별로 쉽게 관찰할 수 있다. 또한, 시뮬레이터는 역전파 알고리즘의 전 방향, 역 방향 연산을 각각 따로 실행시키거나, 병렬로 실행하게 할 수 있다. 병렬 실행은 입력 자료를 연속적으로 입력받아 시스톨릭 어레이의 모든 처리요소들에서 역전파 알고리즘의 양방향 전파를 동시에 실행시킴으로써 가능하다.