• Journal of the Korean Statistical Society
• /
• v.23 no.2
• /
• pp.367-378
• /
• 1994

Let P and Q be probability measures of a measurable space $(\Omega, F)$, and ${F_n}_{n \geq 1}$ be a sequence of increasing sub $\sigma$-fields which generates F. For each $n \geq 1$, let $P_n$ and $Q_n$ be the restrictions of P and Q to $F_n$, respectively. Under the assumption that $Q_n \ll P_n$ for every $n \geq 1$, a zero-one condition is derived for P and Q to have the dichotomy, i.e., either $Q \ll P$ or $Q \perp P$. Then using this condition and the Kolmogorov's zero-one law, we give new and simple proofs of the dichotomy theorems for a pair of Gaussian measures and Poisson processes with examples.

• Bulletin of the Korean Mathematical Society
• /
• v.26 no.1
• /
• pp.83-94
• /
• 1989

• Proceedings of the KSRS Conference
• /
• 2007.10a
• /
• pp.225-228
• /
• 2007

This paper presents a geostatistical methodology to model local uncertainty in spatial estimation of sediment grain size with high-resolution remote sensing imagery. Within a multi-Gaussian framework, the IKONOS imagery is used as local means both to estimate the grain size values and to model local uncertainty at unsample locations. A conditional cumulative distribution function (ccdf) at any locations is defined by mean and variance values which can be estimated by multi-Gaussian kriging with local means. Two ccdf statistics including condition variance and interquartile range are used here as measures of local uncertainty and are compared through a cross validation analysis. In addition to local uncertainty measures, the probabilities of not exceeding or exceeding any grain size value at any locations are retrieved and mapped from the local ccdf models. A case study of Baramarae beach, Korea is carried out to illustrate the potential of geostatistical uncertainty modeling.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.25 no.4
• /
• pp.162-172
• /
• 2021

This paper proposes stochastic methods to find an approximate solution for the L²-Wasserstein least squares problem of Gaussian measures. The variable for the problem is in a set of positive definite matrices. The first proposed stochastic method is a type of classical stochastic gradient methods combined with projection and the second one is a type of variance reduced methods with projection. Their global convergence are analyzed by using the framework of proximal stochastic gradient methods. The convergence of the classical stochastic gradient method combined with projection is established by using diminishing learning rate rule in which the learning rate decreases as the epoch increases but that of the variance reduced method with projection can be established by using constant learning rate. The numerical results show that the present algorithms with a proper learning rate outperforms a gradient projection method.

• Korean Journal of Mathematics
• /
• v.8 no.2
• /
• pp.111-119
• /
• 2000

For a linear operator Q from $R^d$ into $R^d$. ${\alpha}$ > 0 and 0 < $b$ < 1, the Gaussian (Q, $b$, ${\alpha}$)-semi-stability of probability measures on $R^d$ is investigated.

• Proceedings of the Korea Air Pollution Research Association Conference
• /
• 2009.10a
• /
• pp.171-172
• /
• 2009

• Journal of the Korean Mathematical Society
• /
• v.26 no.1
• /
• pp.143-157
• /
• 1989

• Journal of the Korean Data and Information Science Society
• /
• v.24 no.6
• /
• pp.1477-1488
• /
• 2013

In this study, we analyze the dependence structure of KOSPI and NYSE indices based on a two-step estimation procedure. In the rst step, we adopt ARMA-GARCH models with Gaussian mixture innovations for marginal processes. In the second step, time-varying copula parameters are estimated. By using these, we measure the dependence between the two returns with Kendall's tau and Spearman's rho. The two dependence measures for various copulas are illustrated.

• Bulletin of the Korean Mathematical Society
• /
• v.37 no.1
• /
• pp.135-152
• /
• 2000

For a linear operator Q from $R^{d}\; into\; R^{d},\; {\alpha}\;>0\; and\ 0-semi-stability and the operater semi-stability of probability measures on $R^{d}$ are defined. Characterization of $(Q,b,{\alpha})$-semi-stable Gaussian distribution is obtained and the relationship between the class of $(Q,b,{\alpha})$-semi-stable non-Gaussian distributions and that of operator semistable distributions is discussed.

• Journal of the Korean Mathematical Society
• /
• v.56 no.4
• /
• pp.1001-1016
• /
• 2019

We are concerned with optimization methods for the $L^2$-Wasserstein least squares problem of Gaussian measures (alternatively the n-coupling problem). Based on its equivalent form on the convex cone of positive definite matrices of fixed size and the strict convexity of the variance function, we are able to present an implementable (accelerated) gradient method for finding the unique minimizer. Its global convergence rate analysis is provided according to the derived upper bound of Lipschitz constants of the gradient function.