• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권11호
• /
• pp.3913-3934
• /
• 2021

Real-time prediction of Web service of quality (QoS) provides more convenience for web services in cloud environment, but real-time QoS prediction faces severe challenges, especially under the cold-start situation. Existing literatures of real-time QoS predicting ignore that the QoS of a user/service is related to the QoS of other users/services. For example, users/services belonging to the same group of category will have similar QoS values. All of the methods ignore the group relationship because of the complexity of the model. Based on this, we propose a real-time Matrix Factorization based Clustering model (MFC), which uses category information as a new regularization term of the loss function. Specifically, in order to meet the real-time characteristic of the real-time prediction model, and to minimize the complexity of the model, we first map the QoS values of a large number of users/services to a lower-dimensional space by the PCA method, and then use the K-means algorithm calculates user/service category information, and use the average result to obtain a stable final clustering result. Extensive experiments on real-word datasets demonstrate that MFC outperforms other state-of-the-art prediction algorithms.

• 한국전자통신학회논문지
• /
• 제18권1호
• /
• pp.99-106
• /
• 2023

최근 딥러닝 기술을 활용하여 작물 생산량 예측 연구가 많이 진행되고 있다. 딥러닝 알고리즘은 입력 데이터 세트와 작물 예측 결과에 대한 선형 맵을 구성하는데 어려움이 있다. 또한, 알고리즘 구현은 획득한 속성의 비율에 긍정적으로 의존한다. 심층강화학습을 작물 생산량 예측 응용에 적용한다면 이러한 한계점을 보완할 수 있다. 본 논문은 작물 생산량 예측을 개선하기 위해 DQN, Double DQN 및 Dueling DQN 의 성능을 분석한다. DQN 알고리즘은 과대 평가 문제가 제기되지만, Double DQN은 과대 평가를 줄이고 더 나은 결과를 얻을 수 있다. 본 논문에서 제안된 모델은 거짓 판정을 줄이고 예측 정확도를 높이는 것으로 나타났다.

• 환경영향평가
• /
• 제14권6호
• /
• pp.387-402
• /
• 2005

In this study, we tried to predict the change of future land-cover and relationships between land-cover change and geo-spatial information in the Gongju area by using fuzzy logic operation. Quantitative evaluation of prediction models was carried out using a prediction rate curve using. Based on the analysis of correlations between the geo-spatial information and land-cover change, the class with the highest correlation was extracted. Fuzzy operations were used to predict land-cover change and determine the land-cover prediction maps that were the most suitable. It was predicted that in urban areas, the urban expansion of old and new towns would occur centering on the Gem-river, and that urbanization of areas along the interchange and national roads would also expand. Among agricultural areas, areas adjacent to national roads connected to small tributaries of the Gem-river and neighboring areas would likely experience changes. Most of the forest areas are located in southeast and from this result we can guess why the wide chestnut-tree cultivation complex is located in these areas and the possibility of forest damage is very high. As a result of validation using the prediction rate curve, it was indicated that among fuzzy operators, the maximum fuzzy operator was the most suitable for analyzing land-cover change in urban and agricultural areas. Other fuzzy operators resulted in the similar prediction capabilities. However, in the prediction rate curve of integrated models for land-cover prediction in the forest areas, most fuzzy operators resulted in poorer prediction capabilities. Thus, it is necessary to apply new thematic maps or prediction models in connection with the effective prediction of changes in the forest areas.

• 한국대기환경학회:학술대회논문집
• /
• 한국대기환경학회 2009년도 추계학술대회 논문집
• /
• pp.205-206
• /
• 2009

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2012년도 춘계학술대회 논문집
• /
• pp.564-565
• /
• 2012

• 정보처리학회논문지B
• /
• 제17B권3호
• /
• pp.239-248
• /
• 2010

혈액 세포 영상에서 백혈구는 환자의 건강상태를 파악하는데 중요한 정보를 제공하며, 이를 통해 다양한 질병을 조기에 예측할 수 있다. 따라서 질병의 조기 예측을 위해 혈액세포에서 백혈구의 분리는 매우 중요한 단계이다. 본 논문에서는 중요도 맵과 단계적 영역 병합을 이용하여 혈액 세포 영상에서 백혈구를 자동으로 분할하는 기법을 제안한다. 백혈구 세포 영역은 염색물질에 의해 주변에 비해 두드러진 색상, 질감 정보를 가짐으로 이를 기반으로 중요도 맵(Saliency Map)을 만든다. 이를 통해 세포 영상에서 두드러진 영역을 찾아 sub-image를 분리하고, 각 sub-image에서 mean-shift 알고리즘을 적용하여 영역 클러스터링을 수행한다. Mean-shift 적용 후 얻은 클러스터들에 대해 단계적 영역 병합 알고리즘을 적용하고, 최종적으로 백혈구 핵으로 판단되는 단일 클러스터를 얻을 수 있다. 본 논문에서 제안한 방법은 혈액 세포 영상을 사용하여 테스트한 결과 71%의 핵 검출 정확도를 보였으며, 기존의 다른 알고리즘보다 뛰어난 성능을 나타내었다.

• Journal of Biosystems Engineering
• /
• 제27권3호
• /
• pp.227-234
• /
• 2002

This study was conducted to nondestructively estimate the volumetric information of peach and pit and to visualize the 3D information of internal structure from magnetic resonance imaging(MRI) data. Bruker Biospec 7T spectrometer operating at a proton reosonant frequency of 300 MHz was used for acquisition of MRI data of peach. Image processing algorithms and visualization techniques were implemented by using MATLAB (Mathworks) and Visualization Toolkit(Kitware), respectively. Thresholding algorithm and Kohonen's self organizing map(SOM) were applied to MRI data fur region segmentation. Volumetric information were estimated from segemented images and compared to the actual measurements. The average prediction errors of peach and pit volumes were 4.5%, 26.1%, respectively for the thresholding algorithm. and were 2.1%, 19.9%. respectively for the SOM. Although we couldn't get the statistically meaningful results with the limited number of samples, the average prediction errors were lower when the region segmentation was done by SOM rather than thresholding. The 3D visualization techniques such as isosurface construction and volume rendering were successfully implemented, by which we could nondestructively obtain the useful information of internal structures of peach.

• 한국통신학회논문지
• /
• 제24권9A호
• /
• pp.1411-1418
• /
• 1999

본 논문에서는, 움직임 예측과 신경 회로망을 이용한 고속 움직임 추려하여, 현재 블록의 움직임 벡터를 인적 블록들의 움직임 벡터들로 예측하정 알고리즘을 제안하였다. 움직임 벡터의 공간적 상관성이 높다는 점을 고였다. 학습 시간이 빠르고 2차원 적응적 특성의 KSFM(Kohonen self-organizing feature map) 신경망을 이용하여, 움직임 벡터의 코드북(codebook)을 설계하였다. 2차원 코드북상에서 서로 비슷한 코드벡터들(codevectors)은 가까이 위치하므로, 예측 코드벡터로부터 코드북상에서 점진적으로 움직임을 추정하였다. 모의 실험 결과, 제안한 방법이 적은 계산량으로도 우수한 성능을 나타냄을 확인하였다.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2004년도 Proceedings of ISRS 2004
• /
• pp.264-267
• /
• 2004

In identifying flood vulnerable areas, three methods are generally deployed: the geomorphology method which is based on topographic features; the past evidence method based on observed data of past actual floods; and, prediction of flood areas through hydrologic models. This study aims to improve the prediction model of the geomorphology method through the application of fuzzy method in GIS modeling. The generally used GIS method of superimposing thematic map layers assumes crisp boundaries of the layers, which results in either risk-averse solutions or risk-taking solutions. The introduction of fuzzy concepts to processing of evaluation criteria (DEM, slope, aspect) solves this problem. As the result of applying the fuzzy method to a test site in the west Nak-Dong river, similar flood vulnerable areas were predicted as when using the conventional Boolean criteria. The resulting map, however, showed varying degree of uncertainty of flooding in these areas. This extra information is deemed to be valuable in taking phased actions during flood response, leading to a more effective and timely decision-making.

• Asian-Australasian Journal of Animal Sciences
• /
• 제4권4호
• /
• pp.305-328
• /
• 1991

Pig breeding programs have been very successful in the improvement of animals by the simple expedient of focusing on a few traits of economic importance, particularly growth efficiency and leanness. Further reductions in leanness may become more difficult to achieve, due to reduced genetic variation, and less desirable, due to adverse correlated effects on meat and eating quality. Best linear unbiased prediction (BLUP) of breeding values makes possible the incorporation of data from many sources and increases the value of including traits such as sow performance in the breeding objective. Advances in technology, such as electronic animal identification, electronic feeders, improved ultrasonic scanners and automated data capture at slaughter houses, increase the number of sources of information that can be included in breeding value predictions. Breeding program structures will evolve to reflect these changes and a common structure is likely to be several or many breeding farms genetically linked by A.i., with data collected on a number of traits from many sources and integrated into a single breeding value prediction using BLUP. Future developments will include the production of a porcine gene map which may make it possible to identify genes controlling economically valuable traits, such as those for litter size in the Meishan, and introgress them into nucleus populations. Genes identified from the gene map or from other sources will provide insight into the genetic basis of performance and may provide the raw material from which transgenic programs will channel additional genetic variance into nucleus populations undergoing selection.