• 재활복지공학회논문지
• /
• 제1권1호
• /
• pp.21-28
• /
• 2007

대부분의 보청기에서는 난청인의 청력 손실을 보상하기 위해 음성 증폭 알고리즘과 잡음 및 피드백을 제거하기 위한 알고리즘 등을 사용하고 있고, 음성의 인지율을 높이기 위해서 음성의 대비를 강조하는 알고리즘을 사용한다. 음성 대비 강조 알고리즘 사용 시 과도한 대비증폭에 의해서 포먼트(formant) 사이에서 마스킹 현상이 발생할 수 있다. 마스킹 현상을 확인하기 위해서 6가지의 주관적인 평가와 1가지의 객관적인 평가를 수행하였다. 주관적인 평가로는 순음 검사, 어음 청취 역치 검사, 낱말 분별력 검사, 순음 마스킹 검사, 포먼트 순음 마스킹 검사, 어음 마스킹 검사로 이루어져있고 객관적인 평가를 위해서 LLR(log likelihood ratio)을 도입하였다. 정상 청각 피험자와 난청 피험자의 결과로 미루어 볼 때, 난청 피험자가 마스킹이 잘 일어난다는 것을 볼 수 있었고, 어음을 이용한 마스킹 검사에서는 마스킹에 의해서 난청 피험자에서 어음에 대한 인지율이 떨어지는 결과를 볼 수 있다. 이러한 현상은 포먼트 향상에 의한 왜곡 때문이 아닌 마스킹 때문이다. 그러므로 마스킹 효과의 특성을 조사하여, 보청기 적합 시, 이러한 마스킹 특성을 고려하는 것이 필요하다.

• 한국통신학회논문지
• /
• 제28권10C호
• /
• pp.1007-1012
• /
• 2003

Multilayer perceptrons (MLPs)를 위한 일반적인 BP 알고리즘의 학습 속도를 개선하기 위하여 제약을 갖는 최적화 기술을 제안하고 이를 backpropagation (BP) 알고리즘에 적용한다. 먼저 잡음 제약을 갖는 LMS (noise constrained least mean square : NCLMS) 알고리즘과 영잡음 제약 LMS (ZNCLMS) 알고리즘을 BP 알고리즘에 적용한다. 이러한 알고리즘들은 다음과 같은 가정을 반드시 필요로 하여 알고리즘의 이용에 많은 제약을 갖는다. NCLMS 알고리즘을 이용한 NCBP 알고리즘은 정확한 잡음 전력을 알고 있다고 가정한다. 또한 ZNCLMS 알고리즘을 이용한 ZNCBP 알고리즘은 잡음의 전력을 0으로 가정, 즉 잡음을 무시하고 학습을 진행한다. 본 논문에서는 확장된(augmented) Lagrangian multiplier를 이용하여, 비용함수(cost function)를 변형한다. 이를 통하여 잡음에 대한 가정을 제거하고 ZNCBP와 NCBP 알고리즘을 확장, 일반화하여 적응 오류 제약 BP(adaptive error constrained BP : AECBP) 알고리즘을 유도, 제안한다. 제안한 알고리즘들의 수렴 속도는 일반적인 BP 알고리즘보다 약 30배정도 빠른 학습 속도를 나타내었으며, 일반적인 선형 필터와 거의 같은 수렴속도를 나타내었다.

• 한국정보전자통신기술학회논문지
• /
• 제12권3호
• /
• pp.251-256
• /
• 2019

본 연구에서는 손동작을 인식하기 위하여 밀리미터파 기반 레이더에서 얻어진 손동작의 주파수 반향 특성을 이용하는 알고리즘 및 시스템을 제안하였다. 제안된 시스템은 밀리미터파 아날로그 송수신부, ADC부 및 신호 처리부가 원칩으로 구현된 시스템을 이용하여 데이터를 수집하도록 구성하였고, 제안한 알고리즘은 반사된 주파수 영상의 제르니케모멘트로부터 얻어진 전역 및 지역 디스크립터로 구성된 BoF에서 K-means 클러스터링을 이용하여 코드 워드를 생성하고 SVM을 이용한 손동작 분류를 수행하였다. 수행 결과는 혼동행렬에서 얻어진 정밀도, 민감도 및 정확도를 이용하여 평가였다. 정확도 평가에서는 제안한 방법은 GZM방법 및 LZM방법과 비교하여 성능 평가를 위한 인덱스에서 제안한 방법이 95.6%의 성능을 보였고 비교한 나머지 두 방법은 88.4% 및 84%을 나타내어 제안한 방법이 기존의 두 방법에 비하여 7~8% 성능이 향상되었으며 정밀도 및 민감도에서도 나머지 두 방법에 대하여 향상된 성능을 나타내었다. 제안한 방법은 소형화된 밀리미터파 기반 레이더를 이용하여 동작 인식을 할 수 있는 임베디드 시스템의 응용가능성을 보여준다.

• 천문학회보
• /
• 제44권2호
• /
• pp.63.3-63.3
• /
• 2019

Nearby young moving groups (NYMGs hereafter) are gravitationally unbound loose young stellar associations located within 100 pc of the Sun. Since NYMGs are crucial laboratories for studying low-mass stars and planets, intensive searches for NYMG members have been performed. For identification of NYMG members, various strategies and methods have been applied. As a result, the reliability of the members in terms of membership is not uniform, which means that a careful membership re-assessment is required. In this study, I developed a NYMG membership probability calculation tool based on Bayesian inference (Bayesian Assessment of Moving Groups: BAMG). For the development of the BAMG tool, I constructed ellipsoidal models for nine NYMGs via iterative and self-consistent processes. Using BAMG, memberships of claimed members in the literature (N~2000) were evaluated, and 35 per cent of members were confirmed as bona fide members of NYMGs. Based on the deficiency of low-mass members appeared in mass function using these bona fide members, low mass members from Gaia DR2 are identified. About 2000 new M dwarf and brown dwarf candidate members were identified. Memberships of ~70 members with RV from Gaia were confirmed, and the additional ~20 members were confirmed via spectroscopic observation. Not relying on previous knowledge about the existence of nine NYMGs, unsupervised machine learning analyses were applied to NYMG members. K-means and Agglomerative Clustering algorithms result in similar trends of grouping. As a result, six previously known groups (TWA, beta-Pic, Carina, Argus, AB Doradus, and Volans-Carina) were rediscovered. Three the other known groups are recognized as well; however, they are combined into two new separate groups (ThOr+Columba and TucHor+Columba).

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2004년도 ICCAS
• /
• pp.1865-1869
• /
• 2004

A new model for the propagation prediction for mobile communication network inside building is presented in this paper. The model is based on the determination of the dominant paths between the transmitter and the receiver. The field strength is predicted with adaptive neuro - fuzzy inference systems (ANFIS), trained with measurements. The advantage of the ANFIS with hybrid least squares and gradient descent algorithms is fast convergence compared with original neural network. The K-means algorithm for selection of training patterns is also used. Comparison of our predicted results to measurements indicate that improvements in accuracy over conventional empirical model are achieved.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1999년도 추계학술대회 논문집 학회본부 B
• /
• pp.797-799
• /
• 1999

In this paper, we design the Multi-FNN(Fuzzy-Neural Networks) using HCM Method. The proposed Multi-FNN uses simplified inference as fuzzy inference method and Error Back Propagation Algorithm as learning rules. Also, We use HCM(Hard C-Means) method of clustering technique for improvement of output performance from pre-processing of input data. The parameters such as apexes of membership function, learning rates and momentum coefficients are adjusted using genetic algorithms. We use the training and testing data set to obtain a balance between the approximation and the generalization of our model. Several numerical examples are used to evaluate the performance of the our model. From the results, we can obtain higher accuracy and feasibility than any other works presented previously.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1987년도 정기총회 및 창립40주년기념 학술대회 학회본부
• /
• pp.76-78
• /
• 1987

This paper describes the theory and application of a multi-input/multi-output self-tuning regulator where the control objective is the assignment of the closed-loop pole set to prespecified locations. The algorithm described In this paper has a 'self-tuning' property. This self-tuners are more robust than the tuners that are based on optimal control synthesis method. This paper demonstrate usefulness of the algorithms by means of some simulation studies.

• 산업경영시스템학회지
• /
• 제22권53호
• /
• pp.37-44
• /
• 1999

The problem to be studied here is the minimization of the total travel distance and the number of vehicles used for delivering goods to customers. Vehicle routes must also satisfy a variety of constraints such as fixed vehicle capacity, allowed operating time. Genetic algorithm to solve the VRPTW with heterogeneous fleet is presented. The chromosome of the proposed GA in this study has the 3-dimension. We propose GA that has the cubic-chromosome for VRPTW with heterogeneous fleet. The newly suggested ‘Cubic-GA (or 3-D GA)’ in this paper means the 2-D GA with GLS(Genetic Local Search) algorithms and is quite flexible. To evaluate the performance of the algorithm, we apply it to the Solomon's VRPTW instances. It produces a set of good routes and the reasonable number of vehicles.

• 한국정보처리학회논문지
• /
• 제6권3호
• /
• pp.571-579
• /
• 1999

A Macro-Star graph which has a star graph as a basic module has node symmetry, maximum fault tolerance, and hierarchical decomposition property. And, it is an interconnection network which improves a network cost against a star graph. A matrix star graph also has such good properties of a Macro-Star graph and is an interconnection network which has a lower network cost than a Maco-Star graph. In this paper, we propose a method to embed between a Macro-Star graph and a matrix star graph. We show that a Macro-Star graph MS(k, n) can be embedded into a matrix star graph MS\ulcorner with dilation 2. In addition, we show that a matrix star graph MS\ulcorner can be embedded into a Macro-Star graph MS(k,n+1) with dilation 4 and average dilation 3 or less as well. This result means that several algorithms developed in a star graph can be simulated in a matrix star graph with constant cost.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2004년도 ICCAS
• /
• pp.185-189
• /
• 2004

For face recognition system, a face detector which can find exact face region from complex image is needed. Many face detection algorithms have been developed under the assumption that background of the source image is quite simple . this means that face region occupy more than a quarter of the area of the source image or the background is one-colored. Color-based face detection is fast but can't be applicable to the images of which the background color is similar to face color. And the algorithm using neural network needs so many non-face data for training and doesn't guarantee general performance. In this paper, A multi-scale, multi-face detection algorithm using PCA is suggested. This algorithm can find most multi-scaled faces contained in static images with small number of training data in reasonable time.