• Proceedings of the Korea Information Processing Society Conference
• /
• 2009.11a
• /
• pp.273-274
• /
• 2009

본 논문은 완전히 잘못된 데이터가 포함된 회귀(regression) 문제에 사용되는 가우스-균일 혼합확률분포의 두 개의 매개변수 추정에 관하여 고찰한다. 논문에서는 기대값 최대화(Expectation Maximization)와 최우도추정(Maximum Likelihood Estimation)을 이용한 매개변수 추정 방법을 비교한다. 두 기법은 최적화 문제로 기술할 수 있고, 논문에서는 두 기법에서 사용하는 매개변수에 대한 적합도 척도의 개형을 도시하고 비교한다. 몬테-카를로(Monte Carlo) 접근을 통한 두 기법이 추정한 매개변수의 분포를 살펴본다.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1998.06e
• /
• pp.243-246
• /
• 1998

집속음장의 고조파성분을 이용한 초음파영상의 특성을 해석하기 위해 집속된 가우스 음원에 직선 edge를 초점면 및 초점면의 전, 후방에 삽입하여 edge의 후방에서 생성되는 음장을 조사하였다. 계산에서는 그린함수의 간단화를 위해 Fresnel근사를 이용하였고, 실험에서는 성형전극을 형성시킨 1.9MHz 요면진동자에 의한 가우스분포의 음장을 갖는 초음파빔에 수직하게 edge를 삽입시켰다. 음장의 이론해석 및 실험결과, 초점면의 제2고조파의 빔형상을 제외하고는 계산치와 실험치가 잘 일치하고 있으며, 제2고조파의 공간 분해능이 기본파에 비해 높음을 알았다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.6
• /
• pp.1260-1265
• /
• 2012

This paper have presented the change for subthreshold characteristics for double gate(DG) MOSFET based on scaling theory and the shape of Gaussian function. To obtain the analytical solution of Poisson's equation, Gaussian function been used as carrier distribution and consequently potential distributions have been analyzed closely for experimental results, and the subthreshold characteristics have been analyzed for the shape parameters of Gaussian function such as projected range and standard projected deviation. Since this potential model has been verified in the previous papers, we have used this model to analyze the subthreshold chatacteristics. The scaling theory is to sustain constant outputs for the change of device parameters. As a result to apply the scaling theory for DGMOSFET, we know the subthreshold characteristics have been greatly changed, and the change of threshold voltage is bigger relatively.

• Journal of KIISE
• /
• v.42 no.6
• /
• pp.748-754
• /
• 2015

This paper proposes a sequential state estimation model consisting of continuous and discrete variables, as a way of generalizing all discrete-state factorial HMM, and gives a design of gait motion model based on the idea. The discrete state variable implements a Markov chain that models the gait dynamics, and for each state of the Markov chain, we created a Gaussian process over the space of the continuous variable. The Markov chain controls the switching among Gaussian processes, each of which models the rotation or various views of a gait state. Then a particle filter-based algorithm is presented to give an approximate filtering solution. Given an input vector sequence presented over time, this finds a trajectory that follows a Gaussian process and occasionally switches to another dynamically. Experimental results show that the proposed model can provide a very intuitive interpretation of video-based gait into a sequence of poses and a sequence of posture states.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.25 no.1
• /
• pp.11-19
• /
• 2013

As the temperature data show a distribution pattern with a number of peaks, assumption of normal distribution will result in a serious bias in the analysis. In this study, the Gaussian Mixture Distribution (GMD), a type of bimodal distribution, is presumed as a frequency distribution for the water temperature, in order to estimate the optimal parameter and to carry out the relation analysis between the optimal parameter and the basic statistical information such as mean and variance on the data. By the way, an estimation formulae to compute the frequency distribution of the data is developed by computing the parameters of GMD (i.e. ${\alpha}_1$, ${\mu}_1$, ${\sigma}_1$, ${\alpha}_2$, ${\mu}_2$, ${\sigma}_2$) by means of the major characteristic values, such as mean, standard deviation and skewness of the data. The formulae shows an excellent coincidence with the result from the observation data, in the RMS limit accuracy of 5%.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.3
• /
• pp.651-656
• /
• 2014

This paper analyzed the change of subthreshold swing for channel doping of asymmetric double gate(DG) MOSFET. The subthreshold swing is the factor to describe the decreasing rate of off current in the subthreshold region, and plays a very important role in application of digital circuits. Poisson's equation was used to analyze the subthreshold swing for asymmetric DGMOSFET. Asymmetric DGMOSFET could be fabricated with the different top and bottom gate oxide thickness and bias voltage unlike symmetric DGMOSFET. It is investigated in this paper how the doping in channel, gate oxide thickness and gate bias voltages for asymmetric DGMOSFET influenced on subthreshold swing. Gaussian function had been used as doping distribution in solving the Poisson's equation, and the change of subthreshold swing was observed for projected range and standard projected deviation used as parameters of Gaussian distribution. Resultly, the subthreshold swing was greatly changed for doping concentration and profiles, and gate oxide thickness and bias voltage had a big impact on subthreshold swing.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.45 no.6
• /
• pp.57-63
• /
• 2008

In this paper, we present an efficient noise reduction method using bivariate Gaussian density function in the wavelet domain. In our method, the probability model for the interstate dependency in the wavelet domain is modeled by bivariate Gaussian function, and then, the noise reduction is performed by Bayesian estimation. The statistical parameter for Bayesian estimation can be approximately obtained by the $H{\ddot{o}}lder$ inequality. The simulation results show that our method outperforms the previous methods using bivariate probability models.

• The Korean Journal of Applied Statistics
• /
• v.19 no.2
• /
• pp.241-256
• /
• 2006

We generalize the $Cram{\acute{e}}r$-von Mises Statistic to test multivariate normality using Roy's union-intersection principle. We show the limit distribution of the suggested statistic is representable as the integral of a suitable Gaussian process. We also consider the computational aspects of the proposed statistic. Power performance is assessed in a Monte Carlo study.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.10a
• /
• pp.765-768
• /
• 2014

본 연구에서는 비대칭 이중게이트 MOSFET의 상하단 게이트 산화막 두께 비에 대한 문턱전압 및 전도중심의 변화에 대하여 분석하고자한다. 비대칭 이중게이트 MOSFET는 상하단 게이트 산화막의 두께를 다르게 제작할 수 있어 문턱전압이하 영역에서 전류를 제어할 수 있는 요소가 증가하는 장점이 있다. 상하단 게이트 산화막 두께 비에 대한 문턱전압 및 전도중심을 분석하기 위하여 포아송방정식을 이용하여 해석학적 전위분포를 구하였다. 이때 전하분포는 가우스분포함수를 이용하였다. 하단게이트 전압, 채널길이, 채널두께, 이온주입범위 및 분포편차를 파라미터로 하여 문턱전압 및 전도중심의 변화를 관찰한 결과, 문턱전압은 상하단 게이트 산화막 두께 비에 따라 큰 변화를 나타냈다. 특히 채널길이 및 채널두께의 절대값보다 비에 따라 문턱전압이 변하였으며 전도중심이 상단 게이트로 이동할 때 문턱전압은 증가하였다. 또한 분포편차보단 이온주입범위에 따라 문턱전압 및 전도중심이 크게 변화하였다.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.6
• /
• pp.59-68
• /
• 2017

Random process plays a major role in wireless communication system to analytically derive the probability distribution function of the various statistical distribution. In this paper, we derive the decreasing function of the exponential distribution under the given condition which is expressed as wireless channel condition. The probability distribution function of Gaussian, Laplacian, Rayleigh and Nakagami distribution are also derived. Extensive simulation results of these statistical distributions are provided to prove that random process has a significant role in the wireless communications. In addition, the Rayleigh and Rician channels show specific examples of visible distance communication and invisible distance channel environment. This paper is motivated by that we assume a block fading channel model, where the channel is constant during a transmission block and changes independently between consecutive transmission block, can achieve a better performance in high SNR regime with i.i.d channel. This algorithm for realizing these transforms can be applied to the Kronecker MIMO channel.