• 정보처리학회논문지D
• /
• 제11D권2호
• /
• pp.247-258
• /
• 2004

최근 비디오 데이타베이스, 시공간 데이타베이스, 모바일 데이타베이스와 같은 데이타베이스 응용 분야에서 이동 객체를 기반으로 하는 검색 기법에 관한 연구가 활발히 이루어지고 있다. 본 논문에서는 이동 객체의 궤적에 대한 효율적인 유사 부분궤적 검색을 지원하는 새로운 시그니쳐-기반 색인 기법을 제안한다. 제안하는 시그니쳐-기반 색인 기법은 궤적 데이타를 토대로 궤적 시그니쳐를 생성하는 방법에 따라 중첩 시그니쳐-기반 색인 기법(Superimposed signature-based Indexing scheme for similar Sub-trajectory Retrieval : SISR)과 합성 시그니쳐-기반색인 기법(Concatenated signature-based Indexing scheme for similar Sub-trajectory Retrieval : CISR)으로 나뉜다. 생성된 궤적 시그니쳐 정보는 시그니쳐 파일에 저장되고, 검색시 주어진 사용자 질의 궤적 정보를 기반으로 데이타 파일을 직접 접근하기 전에 전체 궤적 시그니쳐들을 탐색하여 필터링을 수행한다. 이를 통해 데이타 파일의 검색 범위를 현저히 줄임으로써 검색 성능을 향상시킨다. 또한 검색된 궤적 데이터와의 유사성을 측정하기 위해 k-워핑 알고리즘을 적용시켜 검색의 효율성을 높인다. 마지막으로, 순차 색인 기법, SISR기법, 그리고 CISR 기법을 삽입시간, 검색 시간 그리고 부가 저장 공간측면에서 성능 평가를 수행한다. 성능 평가 결과, 제안하는 두 가지 기법이 검색 성능 측면에서 순차 색인 기법에 비해 성능이 우수함을 나타내고, 아울러 SISR 기법이 CISR 기법에 비해 보다 우수한 성능을 보인다.

• 정보저장시스템학회:학술대회논문집
• /
• 정보저장시스템학회 2005년도 추계학술대회 논문집
• /
• pp.145-149
• /
• 2005

We describe holographic ROM system to read bit-type data. It has optical system similar to general optical data storage system such as DVD. But because holographic data storage systems have to adopt imaging optical system, in our system bit-type data can be read out by different servos with DVD. We devised 3-hole method similar to 3- beam method for the tracking servo and used astigmatic optical system for the focusing servo. Also we developed the reference beam servo to measure movement of reference beam because especially holographic data storage systems need reference beam. The system was operated by these three servos and objective lens of NA 0.6. We obtained eye pattern from random data of 3T-2um track pitch. We also obtained another eye pattern from DVD disk by only using focusing servo PDIC in our system to verify the compatibility with DVD.

• 한국콘텐츠학회논문지
• /
• 제21권8호
• /
• pp.1-9
• /
• 2021

딥 러닝을 활용한 컴퓨터 비전 연구는 여전히 대규모의 학습 데이터와 컴퓨팅 파워가 필수적이며, 최적의 네트워크 구조를 도출하기 위해 많은 시행착오가 수반된다. 본 연구에서는 네트워크 최적화나 데이터를 보강하는 것과 무관하게 데이터 자체의 특성만을 고려한 CR(Confusion Rate)기반의 유사 이미지 분류 성능 향상 기법을 제안한다. 제안 방법은 유사한 이미지 데이터를 정확히 분류하기 위해 CR을 산출하고 이를 손실 함수의 가중치에 반영함으로서 딥 러닝 모델의 성능을 향상시키는 기법을 제안한다. 제안 방법은 네트워크 최적화 결과와 독립적으로 이미지 분류 성능의 향상을 가져올 수 있으며, 클래스 간의 유사성을 고려해 유사도가 높은 이미지 식별에 적합하다. 제안 방법의 평가결과 HanDB에서는 0.22%, Animal-10N에서는 3.38%의 성능향상을 보였다. 제안한 방법은 다양한 Noisy Labeled 데이터를 활용한 인공지능 연구에 기반이 될 것을 기대한다.

• 대한한의학방제학회지
• /
• 제21권2호
• /
• pp.121-132
• /
• 2013

Objectives : The objectives of this study is to establish data-base to find out similar herbal formulas with a particular herbal formula by comparing composition ratio of configuring herbs. And this thesis is to analyze differences of prescriptions and find out similar prescriptions by utilizing galenical mass ratio, which is directly related to effectiveness of galenical. Methods : This study was proceeded by using Access 2007 with Window 7(MS) and 2,787 prescriptions of which herbal configuration could be indicated by weight unit were analysed from Donguibogam. We standardize all units of the prescription and input the mass ratio data when entered galenical data. Results : We could confirm a degree of similarity between compared prescriptions and a particular prescription according to the sum of differences of herb weight ratio and similarity ratio. Conclusions : A most similar herbal formula could be searched through comparing multi prescriptions by multi prescriptions of herbal configuration from established herbal formula data-base where herb weight ratio of prescriptions is to be input.

• KSBB Journal
• /
• 제27권1호
• /
• pp.40-44
• /
• 2012

The laboratory animal group separation is dividing animal population into subgroups, which have similar average and standard deviation values among the subgroups, based on the biological characteristics such as body weight, glucose level in blood, etc. Although group separation is very important and initial step in experimental design, it needs a labor intensive process for researchers because of making similar average and standard deviation values among the subgroups using the raw biological characteristics. To reduce the labor cost and increase the efficiency of animal grouping, we developed a web program named as laboratory animal group separation (LAGS) program. This LAGS uses biological characteristics of population, number of group, and the number of elements per each subgroup as input data. The LAGS automatically separates the population into each subgroup that has similar statistical data such as average and standard deviation values among subgroups. It also provides researchers with the extraordinary data generated in the process of grouping and the final grouping results by graphical display. Through our LAGS, researchers can validate and confirm results of laboratory animal group separation by just a few mouse clicks.

• 정보처리학회논문지C
• /
• 제10C권7호
• /
• pp.899-914
• /
• 2003

최근의 통신 네트워크에서 teletraffic의 양상은 Poisson 프로세스보다 self-similar프로세스에 의해서 더 잘 반영된다. 이는 통신 네트워크의 teletraffic에 관련하여 self-similar한 성질을 고려하지 않는다면, 통신 네트워크의 성능에 관한 결과는 부정확 할 수밖에 없다는 의미가 된다. 따라서, 통신 네트워크에 관한 시뮬레이션을 수행하기 위한 매우 중요한 요소 중에 하나는 충분히 긴 self-similar한 sequence를 얼마나 잘 생성하느냐의 문제이다. 본 논문에서는 FFT〔20〕, RMD〔12〕 그리고 SRA〔5, 10〕 방법을 이용한 세 개의 pseudo-random self-similar sequence 생성기를 비교 분석하였다. 본 Pseudo-random self-similar sequence 생성기의 성질을 매우 긴 sequence를 생성하는데 요구되는 통계적인 정확도와 생성시간에 대해서 분석하였다. 세 개의 pseudo-random self-similar sequence 생성기의 성능은 Hurst 변수의 상대적인 정확도로 보았을 때는 유사했으나, RMD와 SRA 방법을 이용한 pseudo-random self-similar sequence 생성기가 FFT 방법을 이용한 것보다 속도 면에서는 훨씬 빠른 것으로 나타났다. 또한 본 연구를 통해서 pseudo-random self-similar sequence 생성기의 비교분석을 위한 좀더 좋은 방법이 필요하다는 것을 보여주었다.

• Design & Manufacturing
• /
• 제15권4호
• /
• pp.24-31
• /
• 2021

In this study, an artificial neural network model was constructed to convert CAE analysis data into similar experimental data. In the analysis and experiment, the injection molding data for 50 conditions were acquired through the design of experiment and random selection method. The injection molding conditions and the weight, height, and diameter of the product derived from CAE results were used as the input parameters for learning of the convert model. Also the product qualities of experimental results were used as the output parameters for learning of the convert model. The accuracy of the convert model showed RMSE values of 0.06g, 0.03mm, and 0.03mm in weight, height, and diameter, respectively. As the next step, additional randomly selected conditions were created and CAE analysis was performed. Then, the additional CAE analysis data were converted to similar experimental data through the conversion model. An artificial neural network model was constructed to predict the quality of injection molded product by using converted similar experimental data and injection molding experiment data. The injection molding conditions were used as input parameters for learning of the predicted model and weight, height, and diameter of the product were used as output parameters for learning. As a result of evaluating the performance of the prediction model, the predicted weight, height, and diameter showed RMSE values of 0.11g, 0.03mm, and 0.05mm and in terms of quality criteria of the target product, all of them showed accurate results satisfying the criteria range.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제5권11호
• /
• pp.1946-1958
• /
• 2011

Wireless Sensor Network (WSN) is made up of a number of sensor nodes and base stations. Traffic flow in WSN appears self-similar due to its data delivery process, and this impacts queue length greatly and makes queuing delay worse. Active queue management can be designed to improve QoS performance for WSN. In this paper, we propose self-similar traffic rate estimating algorithm named Power-Law Moving Averaging (PLMA) to regulate packet marking probability. This algorithm improves the availability of the rate estimation algorithm under the self-similar traffic condition. Then, we propose an adaptive Proportional Integral algorithm (SSPI) based on the estimation of the Self-Similar traffic rate by PLMA. Simulation results show that SSPI can achieve lower queue length jitter and smaller setting time than PI.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2006년도 Proceedings of ISRS 2006 PORSEC Volume II
• /
• pp.812-815
• /
• 2006

Proteins are essential agents for controlling, effecting and modulating cellular functions, and proteins with similar sequences have diverged from a common ancestral gene, and have similar structures and functions. Function prediction of unknown proteins remains one of the most challenging problems in bioinformatics. Recently, various computational approaches have been developed for identification of short sequences that are conserved within a family of closely related protein sequence. Protein function is often correlated with highly conserved motifs. Motif is the smallest unit of protein structure and function, and intends to make core part among protein structural and functional components. Therefore, prediction methods using data mining or machine learning have been developed. In this paper, we describe an approach for protein function prediction of motif-based models using data mining. Our work consists of three phrases. We make training and test data set and construct classifier using a training set. Also, through experiments, we evaluate our classifier with other classifiers in point of the accuracy of resulting classification.

• 산업공학
• /
• 제19권4호
• /
• pp.342-354
• /
• 2006

The purpose of this thesis is to develop the schemes of supporting estimate for marine engines’ factories which are in a general make-to-order style. The marine engines’ factories currently use the method which depends on the past data and experiences handled by the responsible person, which causes inefficiency and inaccuracy in dealing with a huge amount of data. We apply association rule to solving the problems mentioned above. Critical data for analysis is filtered among materials that have been using actual records of performance so far. Secondly, relation with each part of marine engines through filtered data so that the company can estimate cost promptly and precisely if customers with similar components as requested. By proposed method of study estimate support efficient and supported exactly.