• Journal of the Korean Data and Information Science Society
• /
• v.20 no.3
• /
• pp.505-514
• /
• 2009

The researcher analyzes the relationship between the number of neighbors and the prediction accuracy in the preference prediction process using collaborative filtering system. The number of neighbors who are involved in the preference prediction process are divided into four groups. Each group shows a little difference in the preference prediction. By using prediction error averages in each group, linear functions are suggested. Through the result of this study, the accuracy of preference prediction can be raised when using linear functions by using the number of neighbors in the suggested system.

• Water for future
• /
• v.33 no.5
• /
• pp.49-55
• /
• 2000

전통적인 매개변수적 목적함수 추정방법은 관측자료의 모든 영역에 걸쳐 선형 또는 지수함수 형태의 가정을 기본으로 매개변수를 추정하는 반면 비매개 변수적 Kernel 가중함수를 이용한 방법은 목적함수의 형태에 대한 가정이 필요 없이 관심 있는 임의의 추정지점에서 이웃하는 자료를 이용하여 목적함수를 국지적으로 근사하는 방법이다. 추계학적 수문학의 전형적인 문제인 "목적함수의 가정"에 의해 발생되는 문제를 줄이려는 노력의 일환으로 비매개변수적 Kernel 가중함수를 이용하는 방법에 연구되었고, 본 지면에서는 Kernel 가중함수를 이용한 비매개변수적 확률밀도함수의 기본이론과 빈도해석, 회귀모형 및 비동질성 천이확률 등의 수문학적 응용에 대하여 살펴보았다.

• 한국IT서비스학회:학술대회논문집
• /
• 2009.05a
• /
• pp.349-352
• /
• 2009

본 연구는 협력적 필터링 기법을 이용한 선호도 예측 과정에서 발생하는 추가 정보를 이용하여 선호도 예측 정확도를 향상시킬 수 있는 방안에 대하여 연구하였다. 본 연구에서는 특정 상품에 대한 목표 고객의 선호도 예측에 선정된 이웃의 수와 선호도 예측 정확도와의 관계를 분석하였다. 분석을 위하여 선호도 예측 과정에 선정된 이웃의 수를 4분위수로 4집단으로 구분하여 구분 집단 간 선호도 예측 정확도에 차이가 나타남을 알 수 있었으며 각 집단의 예측 오차들의 평균들을 이용하여 선형의 보정함수를 제안한다. 본 연구의 결과를 바탕으로 추천시스템에서 이웃 수를 이용한 보정함수를 이용하면 예측 정확도를 높일 수 있다.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.16 no.10
• /
• pp.1816-1823
• /
• 1992

In order to develop general CMAC training technique applicable to any CMAC, characteristics of CMAC learning algorithm and training problems of CMAC are studied. Neighborhood Sequential Training technique which is general and free fro CMAC learning interference is proposed. The technique is used to generate mathematical functions and found to be effective.

• Journal of KIISE:Software and Applications
• /
• v.34 no.3
• /
• pp.246-254
• /
• 2007

Local search algorithms start from a certain candidate solution and probe its neighborhood to find ones with improved quality. This paper proposes a method of probabilistically filtering out bad-looking neighbors based on a simple low-cost preliminary evaluation heuristics. The probabilistic filtering enables us to save time wasted on fully evaluating those solutions that will eventually be trashed, and thus improves the search efficiency by allowing us to spend more time on examining better looking solutions. Experiments with two large-scaled real-world problems, which are a traffic signal control problem in traffic network and a load balancing problem in production scheduling, have shown that the proposed method finds better quality solutions, given the same amount of CPU time.

• Proceedings of the Korean Information Science Society Conference
• /
• 2008.06c
• /
• pp.376-380
• /
• 2008

사회의 복잡화와 인터넷의 성장으로 폭발적으로 늘어나고 있는 정보들을 사용자가 모두 검토한 후 여과하기는 어려운 일이다. 이러한 문제를 보완하기 위해서 자동화된 정보 여과 기술이 사용되는데, k-최근접 이웃(k-nearest neighbor) 알고리즘은 그 구현이 간단하며 비교적 정확하여 가장 널리 쓰이고 있는 알고리즘 중 하나이다. k 개의 최근접 이웃들로부터 평가값을 계산하는 데 흔히 쓰이는 방법은 상관계수를 이용한 가중치에 기반하는 것이다. 본 논문에서는 이를 보완하여 대규모 데이터에 대해서도 속도는 크게 저하되지 않으며 정확도는 대폭 향상시킬 수 있는 방법을 적용하였다. 또한, 최근접 이웃을 구하는 거리함수로 다양한 방법을 시도하였다. 영화추천을 위한 실제 데이터에 대한 실험 결과, 속도의 저하는 미미하였으나 정확도에 있어서는 크게 향상된 결과를 가져올 수 있었다.

• Journal of KIISE:Software and Applications
• /
• v.31 no.6
• /
• pp.829-839
• /
• 2004

Methods based on integer programming have been shown to be very effective in solving various crew pairing optimization problems. However, their applicability is limited to problems with linear constraints and objective functions. Also, those methods often require an unacceptable amount of time and/or memory resources given problems of larger scale. Heuristic methods such as neighborhood search, on the other hand, can handle large-scaled problems without too much difficulty and can be applied to problems having any form of objective functions and constraints. However, neighborhood search often gets stuck at local optima when faced with complex search spaces. This paper presents ,i hybrid algorithm of neighborhood search and integer programming, which nicely combines the advantages of both methods. The hybrid algorithm has been successfully tested on a large-scaled crew pairing optimization problem for a real subway line.

• Journal of the Korea Society of Computer and Information
• /
• v.7 no.4
• /
• pp.204-210
• /
• 2002

The self-organizing feature map of Kohonen has disadvantage that needs too much input patterns in order to converge into the equilibrium state when it trains. In this paper we proposed the method of improving the convergence speed and rate of self-organizing feature map converting the interaction set into Dynamic Gaussian function. The proposed method Provides us with dynamic Properties that the deviation and width of Gaussian function used as an interaction function are narrowed in proportion to learning times and learning rates that varies according to topological position from the winner neuron. In this Paper. we proposed the method of improving the convergence rate and the degree of self-organizing feature map.

• Journal of Korea Multimedia Society
• /
• v.9 no.3
• /
• pp.279-286
• /
• 2006

Classification method used in this paper is applied for the first time to Korean traditional music. Among the frequency distribution vectors, average peak value is suggested and proved effective comparing to previous classification success rate. Mean, variance, spectral centroid, average peak value and ZCR are used to classify Korean traditional musical instruments. To achieve Korean traditional instruments automatic classification, Spectral analysis is used. For the spectral domain, Various functions are introduced to extract features from the data files. k-NN classification algorithm is applied to experiments. Taegum, gayagum and violin are classified in accuracy of 94.44% which is higher than previous success rate 87%.

• Journal of Korea Game Society
• /
• v.20 no.2
• /
• pp.91-100
• /
• 2020

Life-and-death problem in Go is a fundamental problem to be overcome for implementing a computer Go. To solve it, an important consideration is to find out who surrounds or is surrounded between black and white players. To figure out the boundary between black and white groups, we applied an influence function and a contour tracing algorithm. We found that applying the Moore-neighbor tracing among various contour tracing algorithms can create boundaries, and also suggested the possibility of tremendously reducing the search space of a game tree.