• 대한전자공학회논문지TC
• /
• 제47권2호
• /
• pp.40-46
• /
• 2010

다중 사용자 MIMO (Multiple-Input Multiple-Output) 시스템에서 협력적 전송을 기반으로 한 협력적 PF (Proportional Fair) 스케쥴러를 사용하는 경우 사용자 평균 전송률의 로그 합 성능의 개선이 가능하다. 단일 셀 기반의 기존 PF 알고리듬은 여러 채널 환경에 대하여 그 성능이 평가되어 왔으나 여러 개의 기지국들이 스케쥴링에 참여하여 한 프레임 동안 다수의 사용자를 스케쥴링하는 알고리듬의 제시 및 성능 평가는 많은 연구가 필요한 상태이다. 본 논문에서는 서로 다른 셀에 속하는 인접한 세 섹터에 위치한 기지국들이 사용자들의 채널 정보를 교환하여 다중 사용자에게 자원을 할당하는 협력적 PF 스케쥴러를 분산 다중 사용자 MIMO 시스템에 적용하고 사용자 평균 전송률의 로그 합 성능을 평가한다. 또한 그 성능을 기지국 간의 채널 정보 교환 없이 자원을 할당할 사용자를 각자 선택하는 병렬적 PF 스케쥴러와 자원 할당 시 선택할 수 있는 동시 사용자 그룹의 모든 조합을 검색하는 full-search 협력적 PF 스케쥴러의 평균 전송률의 로그 합 성능과 비교 분석한다. 본 연구에 적용된 협력적 PF 스케쥴러는 하위 사용자 평균 전송률의 로그 합 성능 측면에서 병렬적 PF 스케쥴러보다 우수한 성능을 보인다. 또한 모든 조합을 검색함으로서 가장 큰 평균 전송률의 로그 합 성능을 나타내는 full-search 협력적 PF 스케쥴러의 성능의 대부분을 달성하면서도 연산 복잡도를 크게 감소시킨다.

• 한국식품위생안전성학회지
• /
• 제29권4호
• /
• pp.278-284
• /
• 2014

본 연구는 들깻잎과 들깻잎 생산 환경을 대상으로 S. aureus를 분리하고 MLVA를 이용해 분리된 S. aureus의 오염경로를 분석하였다. 그 결과, 들깻잎에서 S. aureus의 검출빈도는 7.9% (9/114)였고 검출수준은 $0{\sim}2.92{\log}_{10}CFU/g$였으며 2곳의 농가에서 S. aureus가 $2.00{\log}_{10}CFU/g$ 이상으로 검출되었다. S. aureus가 분리된 5농가를 대상으로 들깻잎과 직접 접촉하는 시료에서 S. aureus를 분리하였으며 5 농가 중 3농가의 작업환경 (농가의 포장대, 포장비닐, 작업자 손, 복장 등)에서 S. aureus가 검출되었다. 들깻잎과 생산 환경에서 분리된 S. aureus를 대상으로 MLVA를 수행했을 때 포장대, 작업자의 손, 포장비닐 및 관개용수에서 분리된 균주와 들깻잎에서 분리된 S. aureus가 동일한 패턴을 보여 들깻잎의 오염은 이들 시료와 연관되어 있을 것으로 추정된다. 따라서 MLVA를 이용한 유전자 typing은 농산물이나 식품 중 S. aureus의 오염원을 신속하고 경제적으로 구명하는데 활용할 수 있을 것으로 판단된다.

• 천문학회보
• /
• 제37권2호
• /
• pp.83.2-83.2
• /
• 2012

Type I AGNs are classified by the presence of broad emission lines while Type II AGNs show narrow emission lines only. All-sky surveys such as SDSS provide large AGN samples for statistical studies. However, the AGN samples suffer selection bias due to the incomplete selection criteria. To investigate the missing Type I AGNs in optical spectroscopic surveys, we start with a sample of SDSS Type II AGNs at 0.02 < z < 0.05, using the MPA-JHU SDSS DR7 catalog. We search for the hidden broad $H{\alpha}$ component with both visual inspection and the multi-component spectral decomposition method. Out of 1383 Type II AGNs, we find a total of 62 missing Type I AGNs (~4.5%). The sample has mean black hole mass, log $(M_{BH}/M_{SUN))=6.48{\pm}0.53$, and luminosity, log $(L_{H{\alpha}}/ergs^{-1})=40.52{\pm}0.33$, with Eddington ratio, log $(L_{bol}/L_{Edd})=-1.51{\pm}0.41$. We will describe the sample and present the $M_{BH}-{\sigma}_*$, and $M_{BH}-M_*$ relations of the sample in the context of the BH-galaxy coevolution.

• Journal of Information Processing Systems
• /
• 제18권6호
• /
• pp.794-802
• /
• 2022

The recovery of reasonable depth information from different scenes is a popular topic in the field of computer vision. For generating depth maps with better details, we present an efficacious monocular depth prediction framework with coordinate attention and feature fusion. Specifically, the proposed framework contains attention, multi-scale and feature fusion modules. The attention module improves features based on coordinate attention to enhance the predicted effect, whereas the multi-scale module integrates useful low- and high-level contextual features with higher resolution. Moreover, we developed a feature fusion module to combine the heterogeneous features to generate high-quality depth outputs. We also designed a hybrid loss function that measures prediction errors from the perspective of depth and scale-invariant gradients, which contribute to preserving rich details. We conducted the experiments on public RGBD datasets, and the evaluation results show that the proposed scheme can considerably enhance the accuracy of depth prediction, achieving 0.051 for log10 and 0.992 for δ<1.25³ on the NYUv2 dataset.

• 말소리와 음성과학
• /
• 제6권2호
• /
• pp.29-34
• /
• 2014

In the Vector Taylor Series (VTS)-based noisy speech recognition methods, Hidden Markov Models (HMM) are usually trained with clean speech. However, better performance is expected by training the HMM with noisy speech. In a previous study, we could find that Minimum Mean Square Error (MMSE) estimation of the training noisy speech in the log-spectrum domain produce improved recognition results, but since the proposed algorithm was done in the log-spectrum domain, it could not be used for the HMM adaptation. In this paper, we modify the previous algorithm to derive a novel mathematical relation between test and training noisy speech in the cepstrum domain and the mean and covariance of the Multi-condition TRaining (MTR) trained noisy speech HMM are adapted. In the noisy speech recognition experiments on the Aurora 2 database, the proposed method produced 10.6% of relative improvement in Word Error Rates (WERs) over the MTR method while the previous MMSE estimation of the training noisy speech produced 4.3% of relative improvement, which shows the superiority of the proposed method.

• 한국컴퓨터산업학회논문지
• /
• 제3권3호
• /
• pp.363-368
• /
• 2002

전자상거래시스템의 효율적인 운영과 관리를 위해 더욱 많은 노력이 요구되고 있으며 고객의 요구에 대해 가장 적절한 정보를 제공함으로서 만족을 극대화할 수 있어야 한다. 이를 위해 많은 지능형 에이전트 기술을 사용한 전자상거래 시스템이 도입되고 있다. 본 논문에서는 전자상거래시스템에서 CRM 지원을 위한 사례기반 추론과 웹 로그 분석을 기반으로 한 다중 에이전트 시스템을 제안하였다. 본 논문에서 제안한 시스템은 로그분석을 통해 트래픽과 시스템 정보의 분석과 사례기반 추론을 통해 특정 조건에서 고객의 요구에 대해 적절한 정보를 제공할 수 있다.

• Animal Bioscience
• /
• 제34권4호
• /
• pp.482-488
• /
• 2021

Objective: This study was conducted to estimate effect of single nucleotide polymorphisms (SNP) on the estimated breeding value of Hungarian Grey (HG) bulls and to find markers associated with horn colour. Methods: Genotypes 136 HG animals were determined on Geneseek high-density Bovine SNP 150K BeadChip. A multi-locus mixed-model was applied for statistical analyses. Results: Six SNPs were identified to be associated (-log10P>10) with green and white horn. These loci are located on chromosome 1, 3, 9, 18, and 25. Seven loci (on chromosome 1, 3, 6, 9, 10, 28) showed considerable association (-log10P>10) with the estimated breeding value. Conclusion: Analysis provides markers for further research of horn colour and supplies markers to achieve more effective selection work regarding estimated breeding value of HG.

• Asian Pacific Journal of Cancer Prevention
• /
• 제14권11호
• /
• pp.6751-6755
• /
• 2013

Background: Research on cancers with a high rate of mortality such as those occurring in the stomach requires using models which can provide a closer examination of disease processes and provide researchers with more accurate data. Various models have been designed based on this issue and the present study aimed at evaluating such models. Materials and Methods: Data from 330 patients with gastric cancer undergoing surgery at Iran Cancer Institute from 1995 to 1999 were analyzed. Cox-Snell Residuals and Akaike Information Criterion were used to compare parametric and semi-parametric Cox models in modeling transition rates among different states of a multi-state model. R 2.15.1 software was used for all data analyses. Results: Analysis of Cox-Snell Residuals and Akaike Information Criterion for all probable transitions among different states revealed that parametric models represented a better fitness. Log-logistic, Gompertz and Log-normal models were good choices for modeling transition rate for relapse hazard (state $1{\rightarrow}state$ 2), death hazard without a relapse (state $1{\rightarrow}state$ 3) and death hazard with a relapse (state $2{\rightarrow}state$ 3), respectively. Conclusions: Although the semi-parametric Cox model is often used by most cancer researchers in modeling transition rates of multistate models, parametric models in similar situations- as they do not need proportional hazards assumption and consider a specific statistical distribution for time to occurrence of next state in case this assumption is not made - are more credible alternatives.

• Animal Bioscience
• /
• 제35권3호
• /
• pp.377-384
• /
• 2022

Objective: This study was conducted to estimate multi-trait genetic parameters for somatic cell score (SCS), milk yield and type traits in Nigerian Dwarf (ND) goats from the United States. Methods: Data from 1,041 ND goats in the United States with kiddings in 95 herds were used to estimate multi-trait genetic parameters for SCS, milk (MILK), fat (FAT), and protein (PROT) yields, and 14 type traits. An 18-trait mixed linear animal model for lactation mean SCS (Log₂), MILK, FAT, PROT, and 14 type traits was applied. A factor analytic approach (FA1) in ASReml software was used to obtain convergence. Results: Averages for SCS were low (2.85±1.29 Log₂), and were 314±110.6, 20.9±7.4, and 14±4.9 kg, respectively, for MILK, FAT, and PROT. Heritabilities for SCS, MILK, FAT, and PROT were 0.32, 0.16, 0.16, and 0.10, respectively. The highest heritabilities for type traits were for stature (0.72), teat diameter (0.49), and rump width (0.48), and the lowest estimates were for dairyness (0.003) and medial suspensory ligament (0.03). Genetic correlations of SCS with MILK, FAT, and PROT were positive but low (0.25, 0.18, and 0.23, respectively). Genetic and phenotypic correlations between MILK, FAT, and PROT were high and positive (≥0.66). Absolute values of genetic correlations involving SCS with type traits were generally low or no different from zero. Most of the phenotypic correlations involving SCS with type traits were low. No serious unfavorable genetic correlations between milk yield traits and SCS or between milk yield traits or SCS and type traits were found. Conclusion: Genetic variation exists in the ND breed for most studied traits. The development of selection programs based on these estimates may help accelerate favorable multi-trait genetic changes in this breed.

• Journal of Communications and Networks
• /
• 제18권3호
• /
• pp.333-339
• /
• 2016

A multi-stage turbo equalization scheme based on the bit-level combining (BLC) is proposed for multiple-input multiple-output (MIMO) systems with hybrid automatic repeat request (HARQ). In the proposed multi-stage turbo equalization scheme, the minimum mean-square-error equalizer at each iteration calculates the extrinsic log-likelihood ratios for the transmitted bits in a subpacket and the subpackets are sequentially replaced at each iteration according to the HARQ rounds of received subpackets. Therefore, a number of iterations are executed for different subpackets received at several HARQ rounds, and the transmitted bits received at the previous HARQ rounds as well as the current HARQ round can be estimated from the combined information up to the current HARQ round. In addition, the proposed multi-stage turbo equalization scheme has the same computational complexity as the conventional bit-level combining based turbo equalization scheme. Simulation results show that the proposed multi-stage turbo equalization scheme outperforms the conventional BLC based turbo equalization scheme for MIMO systems with HARQ.