• 제목/요약/키워드: mixture distributions

검색결과 272건 처리시간 0.02초

모의실험을 통한 두 처리군간 치료율 비교방법 연구 (Simulation Study for Statistical Methods in Comparing Cure Rates between Two Groups)

  • 박미라;이재원;진서훈
    • 응용통계연구
    • /
    • 제17권2호
    • /
    • pp.253-267
    • /
    • 2004
  • 임상시험중에는 소아암연구에서와 같이 환자 중의 상당수에서 사망 또는 재발이 오랜 기간 일어나지 않고 완치된 것으로 보이는 경우가 있다. 이 경우 연구자는 생존함수의 전반적인 비교보다는 치료율의 비교에 더 관심이 있을 것이다. 본고에서는 치료율의 비교를 위한 여러 모수적, 비모수적 방법들을 소개하고, 생존분포, 치료율, 중도절단을 등을 다양하게 설정한 모의실험을 통하여 각 방법들의 검정력과 유의수준을 비교하였다.

분포함수를 기초로 일반화가중선형모형 (Generalized Weighted Linear Models Based on Distribution Functions - A Frequentist Perspective)

  • 여인권
    • 응용통계연구
    • /
    • 제17권3호
    • /
    • pp.489-498
    • /
    • 2004
  • 이 논문에서는 일반화가중선형모형이라는 새로운 형태의 선형모형을 제시한다. 일반화가중선형모형은 설명변수와 반응변수의 관계를 설명분포함수의 선형결합이 반응변수의 평균에 대한 연결분포함수를 통해 모형화 되는 형태를 가지는 것으로 가정한다. 이모형은 일반화선형 모형에서 연결함수를 선택할 때 발생할 수 있는 모수공간과 선형 예측값의 공간이 일치하지 않을 수 있다는 문제가 발생하지 않고 모수에 대한 해석이 용이하다는 장점이 있다. 이 논문에서는 설명분포함수와 연결분포함수를 선택하는데 있어 발생할 수 있는 문제와 해결책에 대해 알아본다. 또한 모형에 포함되어 있는 모수를 추정하는데 고려해야 할 주의 사항과 이 사항들을 고려한 최대가능도추정법과 재표집 방법을 이용한 구간추정과 가설검정에 대해 알아본다.

라플라스와 이중 파레토 벌점의 비교: LASSO와 Elastic Net (Comparison of Laplace and Double Pareto Penalty: LASSO and Elastic Net)

  • 경민정
    • 응용통계연구
    • /
    • 제27권6호
    • /
    • pp.975-989
    • /
    • 2014
  • 연속적인 변수 선택과 계수 추정을 동시에 활용할 수 있다는 특성 때문에 LASSO (Tibshirani, 1996)와 Elastic Net (Zou와 Hastie, 2005)은 다양한 분야에서 활발하게 사용되고 있다. 조건부 라플라스와 이중 파레토 사전분포를 적용한 공액계층모형을 표현하였고, 각각의 사전분포에 대한 완전 조건 사후분포를 도출하였다. 제안된 사전분포를 적용한 벌점회귀모형을 비교하기 위한 모의 실험을 진행하였고, 예측정확도를 판단하기 위해 아시아 국가 실패(the collapse of governments in Asia)의 실제 데이터에 제안한 모형을 적용하였다.

연속 잡음 음성 인식을 위한 다 모델 기반 인식기의 성능 향상에 대한 연구 (Performance Improvement in the Multi-Model Based Speech Recognizer for Continuous Noisy Speech Recognition)

  • 정용주
    • 음성과학
    • /
    • 제15권2호
    • /
    • pp.55-65
    • /
    • 2008
  • Recently, the multi-model based speech recognizer has been used quite successfully for noisy speech recognition. For the selection of the reference HMM (hidden Markov model) which best matches the noise type and SNR (signal to noise ratio) of the input testing speech, the estimation of the SNR value using the VAD (voice activity detection) algorithm and the classification of the noise type based on the GMM (Gaussian mixture model) have been done separately in the multi-model framework. As the SNR estimation process is vulnerable to errors, we propose an efficient method which can classify simultaneously the SNR values and noise types. The KL (Kullback-Leibler) distance between the single Gaussian distributions for the noise signal during the training and testing is utilized for the classification. The recognition experiments have been done on the Aurora 2 database showing the usefulness of the model compensation method in the multi-model based speech recognizer. We could also see that further performance improvement was achievable by combining the probability density function of the MCT (multi-condition training) with that of the reference HMM compensated by the D-JA (data-driven Jacobian adaptation) in the multi-model based speech recognizer.

  • PDF

Buckling analysis of sandwich beam reinforced by GPLs using various shear deformation theories

  • Hanifehlou, Sona;Mohammadimehr, Mehdi
    • Computers and Concrete
    • /
    • 제25권5호
    • /
    • pp.427-432
    • /
    • 2020
  • In this research, the buckling analysis of sandwich beam with composite reinforced by graphene platelets (GPLs) in two face sheets is investigated. Three type various porosity patterns including uniform, symmetric and asymmetric are considered through the thickness direction of the core. Also, the top and bottom face sheets layers are considered composite reinforced by GPLs/CNTs based on Halpin-Tsai micromechanics model and extended mixture rule, respectively. Based on various shear deformation theories such as Euler-Bernoulli, Timoshenko and Reddy beam theories, the governing equations of equilibrium using minimum total potential energy are obtained. It is seen that the critical buckling load decreases with an increase in the porous coefficient, because the stiffness of sandwich beam reduces. Also, it is shown that the critical buckling load for asymmetric distribution is lower than the other cases. It can see that the effect of graphene platelets on the critical buckling load is higher than carbon nanotubes. Moreover, it is seen that the difference between carbon nanotubes and graphene platelets for Reddy and Euler-Bernoulli beam theories is most and least, respectively.

전자 제어식 가솔린 엔진의 벽류 생성 요인에 관한 연구 (A Study on the Factors of Fuel-Film Formation in an EGI Gasoline Engine)

  • 김봉규;이기형;이창식
    • 대한기계학회논문집B
    • /
    • 제22권11호
    • /
    • pp.1530-1537
    • /
    • 1998
  • Mixture formation is one of the significant factors to improve combustion performance of an spark ignition engine. This is affected by spray and atomization characteristics of injector. In the case of EGI system, air-fuel mixing period is so short that a lot of fuel-film and liquid-fuel flow into cylinder. Since this fuel-film is not burnt perfectly in cylinder, it is exhausted in the form of HC emission. In this paper, three measurement techniques were utilized to measure spray characteristics and the amount of fuel-film in the cylinder. At first, PMAS was used to measure the spray characteristics such as size distributions, SMD, and spray angle. Secondly the amount and distribution of fuel-film which flow into through intake valve could be measured quantitatively using the fuel-film measuring device. And lastly, by optical fiber type spark plug used to detect the diffusion flame, the amount of unburned HC was measured. As the result of these experiments, the information of optimal spray characteristics and injection condition to minimize fuel-film could be built up.

Organic Carbon, Calcium Carbonate, and Clay Mineral Distributions in the Korea Strait Region, the Southern Part of the East Sea

  • Khim, Boo-Keun;Shin, Dong-Hyeok;Han, Sang-Joon
    • Journal of the korean society of oceanography
    • /
    • 제32권3호
    • /
    • pp.128-137
    • /
    • 1997
  • This study presents results from a detailed sedimentological investigation of surface sediments obtained from the Korea Strait region, the southern part of the East Sea (Sea of Japan). The distribution of different types of bottom sediments is controlled by the recent fine-grained sediment transport and deposition combined with the lowerings of sea level during the last glacial period, forming a diverse mixture of organic-rich fine-grained and shelly coarse-grained sediments. In comparison to high organic concentration of fine-grained sediments in the inner continental shelf and slope areas, the shell-rich coarse-grained sediments on the outer shelf are discernible being further modified. These coarse-grained sediments are confirmed as relict resulting from the sediment dynamics during the lower sea levels of the last glacial period. Clay mineral distribution of the fine-grained sediments gives information about the transport mechanism. Presence of present-day current system (the Tsushima Warm Current) is most probable source for the fine-grained particles into the open East Sea from the East China Sea, indicating that Holocene sediment dynamics may be used to explain the observed distribution of surface coarse-grained shell-rich sediments.

  • PDF

인천연안 퇴적물 중 PCB의 분포 (The Distributions of PCBs in Inchon Coastal Sediments)

  • 이동훈;김교근
    • 분석과학
    • /
    • 제15권3호
    • /
    • pp.287-299
    • /
    • 2002
  • 인천연안지역 표층퇴적물의 PCBs 잔류농도는 우리 나라의 다른 연안지역 보다 비교적 낮은 농도로 검출되었고, 전지역에 걸쳐 Aroclors 혼합표준액 (1016+1232+1248+1260)과 유사한 패턴으로 배출되었으며, 염소조성비율 변화에 따라 A(Kanghwa), B(Yellow) 및 C(Jangdo) 등 3개의 지역군으로 분류되었다. 또한, 외해에 근접한 B 지역은 2, 3 염화물이, 매립지 침출수 처리수가 방류되는 C지역은 7, 8염화물이 다른 지역에 비해 높은 염소 조성비율을 나타내는 것으로 조사되었다. 연구대상지역 주상퇴적물의 PCBs 잔류농도는 강화수로에서 1.60~3.00 ng/g, 장도수문에서 0.34~9.20 ng/g으로 나타났으며, 두 지역 모두 하부층에서 상부층으로 올라갈수록 농도가 증가하는 경향을 나타내었다.

자탈형(自脱型) 콤바인의 배진손실(排塵損失)에 관(關)한 연구(硏究)(I) -자탈형(自脱型) 콤바인의 배진손실(排塵損失) 모니터 개발- (Chaff-outlet Grain Loss of Head-feed Combine -Development of a Monitor for Detecting Chaff-outlet Grain Loss of Head-feed Combine-)

  • 정창주;최중섭;최영수
    • Journal of Biosystems Engineering
    • /
    • 제14권4호
    • /
    • pp.251-261
    • /
    • 1989
  • The amount of grain loss incurred during harvesting operation through the chaff-outlet of combine may not be negligible. To minimize this grain loss and optimize condition of combining, it may be necessary that the amount of chaff-outlet loss dependent on varying crop condition is to be estimated as exactly as possible. This study was thus intended to develop the monitor that could indicate the amount of grain loss occurred through the chaff-outlet of combine during harvesting operation. The function of this monitor is to measure at the sounding board the impact sound of paddy kernels which could be distinguished from those of other threshing products through chaff-outlet, and from vibration or noise created by the combine engine and other moving parts. To develop such monitor, the frequency distributions of each sound generated by the impact of grain and chaff, the sound generated by the impact of the mixture of grain and chaff, and vibration or noise created by the combine engine and other moving parts were investigated experimentally. From the results of frequency analyses, the trainsducer adequate for the monitering system was selected and sounding board was constructed. The grain loss monitor thus obtained was tested by attaching the sounding board to the chaff-outlet of combine.

  • PDF

용탕단조법에 의한 고강도 Mg-Li-Al합금 제조 (Fabrication of High Strength Mg-Li-Al Alloys by Squeeze Casting Process)

  • 한창화;황영하;김영우;김도향;홍준표
    • 한국주조공학회지
    • /
    • 제17권3호
    • /
    • pp.267-275
    • /
    • 1997
  • Fabrication of high strength Mg-Li-Al alloys by squeeze casting was established by the stabilization of melt and mold temperatures, applied pressure and the refining method. The entrapment of inclusions during pouring was prevented using 30 ppi alumina foam filter. The as-cast microstructure consists of a mixture of ${\alpha}$ and ${\beta}$ phases including AILi and $MgLi_2$, Al particles, which are distributed in the ${\beta}$ matrix. The grain sizes of gravity and squeeze casting alloys were 288 ${\mu}m$ and 207 ${\mu}m$ respectively. The addition of Al in Mg-Li alloys promoted the formation of second phase particles, which were adjusted to optimize the properties of Mg-Li-Al alloys. The Mg-10wt%Li-5wt%Al alloy after heat treatment at $350^{\circ}C$ for 1 hour showed the maximum hardness value. This is due to the facts that the amounts of ${\alpha}$ and ${\beta}$ phases and their distributions are dependent upon the solution treatment temperature, and that the amounts of AILi and $MgLi_2Al$ particles are dependent upon the Al content.

  • PDF