• 한국해양공학회지
• /
• 제33권3호
• /
• pp.222-228
• /
• 2019

An extreme value analysis (EVA) is essential to obtain a design value for highly nonlinear variables such as long-term environmental data for wind and waves, and slamming or sloshing impact pressures. According to the extreme value theory (EVT), the extreme value distribution is derived by multiplying the initial cumulative distribution functions for independent and identically distributed (IID) random variables. However, in the position mooring of DNVGL, the sampled global maxima of the mooring line tension are assumed to be IID stochastic variables without checking their independence. The ITTC Recommended Procedures and Guidelines for Sloshing Model Tests never deal with the independence of the sampling data. Hence, a design value estimated without the IID check would be under- or over-estimated because of considering observations far away from a Weibull or generalized Pareto distribution (GPD) as outliers. In this study, the IID sampling data are first checked in an EVA. With no IID random variables, an automatic resampling scheme is recommended using the block maxima approach for a generalized extreme value (GEV) distribution and peaks-over-threshold (POT) approach for a GPD. A partial autocorrelation function (PACF) is used to check the IID variables. In this study, only one 5 h sample of sloshing test results was used for a feasibility study of the resampling IID variables approach. Based on this study, the resampling IID variables may reduce the number of outliers, and the statistically more appropriate design value could be achieved with independent samples.

• 인터넷정보학회논문지
• /
• 제24권6호
• /
• pp.1-11
• /
• 2023

Federated learning (FL) is a ground breaking machine learning paradigm that allow smultiple participants to collaboratively train models in a cloud environment, all while maintaining the privacy of their raw data. This approach is in valuable in applications involving sensitive or geographically distributed data. However, one of the challenges in FL is dealing with heterogeneous and non-independent and identically distributed (non-IID) data across participants, which can result in suboptimal model performance compared to traditionalmachine learning methods. To tackle this, we introduce FedGCD, a novel FL algorithm that employs Graph Neural Network (GNN)-based community detection to enhance model convergence in federated settings. In our experiments, FedGCD consistently outperformed existing FL algorithms in various scenarios: for instance, in a non-IID environment, it achieved an accuracy of 0.9113, a precision of 0.8798,and an F1-Score of 0.8972. In a semi-IID setting, it demonstrated the highest accuracy at 0.9315 and an impressive F1-Score of 0.9312. We also introduce a new metric, nonIIDness, to quantitatively measure the degree of data heterogeneity. Our results indicate that FedGCD not only addresses the challenges of data heterogeneity and non-IIDness but also sets new benchmarks for FL algorithms. The community detection approach adopted in FedGCD has broader implications, suggesting that it could be adapted for other distributed machine learning scenarios, thereby improving model performance and convergence across a range of applications.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2008년도 하계종합학술대회
• /
• pp.213-214
• /
• 2008

In this paper, we analyze the performance of a cooperative communication wireless network over independent and identically distributed (IID) Nakagami-m fading channels. A simple transmission scheme is considered where the relay is operating in amplify-forward (AF) mode. A closed-form expression for symbol error rate (SER) is obtained using the moment generating function (MGF) of the total signal to noise ratio (SNR) of the transmitted signal with binary phase shift keying (BPSK).

• 한국전자파학회논문지
• /
• 제28권10호
• /
• pp.825-831
• /
• 2017

실제 레이다가 운용되는 환경에서 발생되는 클러터는 비균질성(heterogeneous)의 특성을 갖는 동시에 바이스태틱 레이다나 모노스태틱 non-sidelooking 레이다 구조인 경우는 클러터의 비정상성(nonstationary) 특성도 갖는다. 이러한 특성에 의해서 클러터 신호를 추정하는데 필요한 IID(Independent Identically Distributed) secondary 데이터 개수에 제약이 따르므로 클러터 억제 성능이 저하된다. 본 논문에서는 바이스태틱 레이다 환경에서 Cell under test에 대한 사전정보만을 이용하여 클러터 신호를 추정함으로써 secondary 데이터 없이 클러터를 억제하는 알고리즘을 제시한다. 바이스태틱 클러터의 angle-Doppler 스펙트럼 상에서 구조 분석을 통해 사전정보로 부터 클러터를 추정하는 것이 가능함을 보이고, 고유치 해석에 의해 클러터 억제 과정을 제시한다. 마지막으로 시뮬레이션을 통해 제시하는 클러터 억제 알고리즘의 성능을 보인다.

• 인지과학
• /
• 제10권3호
• /
• pp.39-49
• /
• 1999

본 연구에서 사용된 결 자극(texture stimulus)은 확률의 개념을 이용하여 만들어졌다. 결 자극은 결 요소들의 확률분포로부터 계산된 평균과 분산으로 완전하게 기술되어 질 수 있으며, 또한 평균과 분산이 서로 직교로(orthogonal) 변화할 수 있는 유클리드 공간 속에 위치할 수 있게 된다. 결 분리과정에 관여하는 기제로는 결의 평균정보를 사용하는 기제와 결의 분산정보를 사용하는 기제가 있을 것으로 가정하였다. 본 실험에서는 유클리드 결 공간에서 짝 지워진 결 자극을 분리하는 확률을 결 자극의 평균차이와 분산차이의 함수로서 측정하였다. 두 명의 피험자로부터 얻어진 자료는 평균과 분산으로 정의된 결 공간에 두 가지 기제가 결 분리과정에 관여하고 있음을 보여주었다. 그리고 두 기제의 반응을 확률 총합(probability summation) 원리에 따라 종합적으로 처리한다는 모형이 자료를 잘 설명하였다. 그러나 두 기제가 담당하는 결 자극의 각각의 차원이 결의 평균과 분산은 아닌 것으로 밝혀졌다. 그러므로 평균과 분산이 독립적인 축을 형성하도록 구성된 유클리드 공간에서 생성된 결 자극의 분리과정에는 두 개의 상호 독립적인 기제가 관여하고 있다. 각 기제의 특성을 살펴보면, 한 기제는 결 자극의 밝기 정보를 처리하고, 다른 기제는 결의 밝기(luminance)정보와 대비 성(contrastness) 정보를 동시에 처리하는 것처럼 보인다.

• 대한교통학회지
• /
• 제21권4호
• /
• pp.57-66
• /
• 2003

로짓모형은 선택대안에 대한 확률 계산이 용이하고, 설명변수의 파라메타 추정이 용이하기 때문에 교통 수단 선택모형으로 널리 쓰여지고 있다. 그러나 이러한 로짓모형은 수단선택 효용함수의 오차항 분포가 선택 대안간에 독립적이고, 그 분산이 동일하다는(IID:Independent and Identically Distributed)가정을 내포한다. 본 연구는 수단선택 효용오차의 분산이 수단간에 동일하다는 가정을 완화시키는 이분산 로짓모형 추정에 관한 연구이다. 수단선택 효용오차항의 동분산성을 극복함으로써 보다 현실적인 통행자의 수단선택행태를 반영하는 로짓모형을 추정하는데 본 연구의 목적이 있다. 이를 위해 로짓모형 오차항의 분산과 직접적인 관련이 있는 규모인자(scale factor)를 도입하였다. 이는 대중 교통과 승용차의 통행시간차이에 따른 이분산성을 고려하도록 정의되었으며, 이를 통행시간 파라메타 추정에 활용하였다. 본 연구에서 개발된 이분산 로짓모형의 추정 결과. 통행자의 통행시간이 증가하면서 대중교통수단과 승용차의 통행시간차이가 동일하더라도 통행자의 대중교통 수단선택확률이 차이를 보임으로 현실적인 통행자의 수단선택 행태를 반영하는 것으로 판명되었다.

• 국제학술발표논문집
• /
• The 5th International Conference on Construction Engineering and Project Management
• /
• pp.21-27
• /
• 2013

Construction safety is a predominant hindrance in in-situ workflow and considered an unresolved issue. Current methods used for safety optimization and prediction, with limited exceptions, are paper-based, thus error prone, as well as time and cost ineffective. In an attempt to exploit the potential of BIM for safety, the objective of the proposed methodology is to automatically predict hazardous on-site conditions related to the route that the dozers follow during the different phases of the project. For that purpose, safety routes used by construction equipment from an origin to multiple destinations are computed using video cameras and their cycle times are calculated. The cycle times and factors; including weather and light conditions, are considered to be independent and identically distributed random variables (iid); and simulated using the Arena software. The simulation clock is set to 100 to observe the minor changes occurring due to external parameters. The validation of this technology explores the capabilities of BIM combined with simulation for enhancing productivity and improving safety conditions a-priori. Preliminary results of 262 measurements indicate that the proposed methodology has the potential to predict with 87% the location of exclusion zones. Also, the cycle time is estimated with an accuracy of 89%.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제7권12호
• /
• pp.3118-3134
• /
• 2013

This paper deliberates for a multimedia transmission scheme combining multiple description coding (MDC) and network coding (NC). Our goal is to take advantage from the property of MDC to provide quantized and compressed independent and identically distributed (iid) descriptions and also from the benefit of network coding, which uses network resources efficiently to recover lost data in the network. Recently, p-cycle NC has been introduced to recover and protect any lost or distorted descriptions at the receiver part exactly without need of retransmission. So far, MDC have not been explored using this type of NC. Compressed and coded descriptions are transmitted through the network where p-cycle NC is applied. P-cycle based algorithm is proposed for single and multiple descriptions lost. Results show that in the fixed bit rate, the PSNR (Peak Signal to Noise Ratio) of our reconstructed image and also subjective evaluation is improved significantly compared to previous work which is averaging method joint with MDC in order to conceal lost descriptions.

• 대한안전경영과학회지
• /
• 제12권3호
• /
• pp.257-262
• /
• 2010

The study investigates the various Acceptance Control Charts (ACCs) based on the factors that include process independence, data weighting scheme, subgrouping, and use of control charts. USL - LSL > $6{\sigma}$ that used in the good condition processes in the ACCs are designed by considering user's perspective, producer's perspective and both perspectives. ACCs developed from the research is efficiently applied by using the simple control limit unified with APL (Acceptable Process Level), RLP (Rejectable Process Level), Type I Error $\alpha$, and Type II Error $\beta$. Sampling interval of subgroup examines i.i.d. (Identically and Independent Distributed) or auto-correlated processes. Three types of weight schemes according to the reliability of data include Shewhart, Moving Average(MA) and Exponentially Weighted Moving Average (EWMA) which are considered when designing ACCs. Two types of control charts by the purpose of improvement are also presented. Overall, $\alpha$, $\beta$ and APL for nonconforming proportion and RPL of claim proportion can be designed by practioners who emphasize productivity and claim defense cost.