• 한국지구물리탐사학회:학술대회논문집
• /
• 2005.05a
• /
• pp.227-232
• /
• 2005

Least squares (${\ell}^2-norm$) solutions of seismic inversion tend to be very sensitive to data points with large errors. The ${\ell}^p-norm$ minimization for $1{\le}p<2$ gives more robust solutions, but usually with higher computational cost. Iteratively reweighted least squares (IRLS) gives efficient approximate solutions of these ${\ell}^p-norm$ problems. I propose a simple way to implement the IRLS method for a hybrid ${\ell}^1/{\ell}^2$ minimization problem that behaves as ${\ell}^2$ fit for small residual and ${\ell}^1$ fit for large residuals. Synthetic and a field-data examples demonstrates the improvement of the hybrid method over least squares when there are outliers in the data.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.22 no.3
• /
• pp.312-319
• /
• 2011

In this paper, a new linearization algorithm for DPD(Digital PreDistorter) is suggested. This new algorithm uses DFP(Davidon-Fletcher-Powell) method. This algorithm is more accurate than that of the existing algorithms, and this method renew the best-fit value in every routine with out setting the initial value of step-size. In modeling power amplifier, the memory polynomial model which can model the memory effect of the power amplifier is used. And the overall structure of linearizer is based on an indirect learning architecture. In order to verify for performance of proposed algorithm, we compared with LMS(Least Mean-Squares), RLS(Recursive Least squares) algorithm.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.10 no.3
• /
• pp.143-151
• /
• 2006

This paper presents the comparison of the goodness-of-fit test of analytical bond models between concrete and steel or GFRP reinforcements. Bond test specimens were prepared in accordance with the CSA codes and the rebars used in the test were steel and two types of commercial GFRP rebar products. Using the test data, a bond model was proposed, and comparison of goodness-of-fit test for existing bond models and proposed bond model was carried out by the least squares method. The result indicates that the proposed bond model has better goodness-of-fit test than the existing ones.

• The Pure and Applied Mathematics
• /
• v.11 no.1
• /
• pp.37-49
• /
• 2004

We are interested in the problem of fitting a sphere to a set of data points in the three dimensional Euclidean space. In Spath [6] a descent algorithm already have been given to find the sphere of best fit in least squares sense of minimizing the orthogonal distances to the given data points. In this paper we present another new algorithm which computes a parametric represented sphere in order to minimize the sum of the squares of the distances to the given points. For any choice of initial approximations our algorithm has the advantage of ensuring convergence to a local minimum. Numerical examples are given.

• Geophysics and Geophysical Exploration
• /
• v.10 no.2
• /
• pp.124-130
• /
• 2007

Least squares ($l^2$ norm) solutions of seismic inversion tend to be very sensitive to data points with large errors. The $l^1$ norm minimization gives more robust solutions, but usually with higher computational cost. Iteratively reweighted least squares (IRLS) method gives efficient approximate solutions of these $l^1$ norm problems. I propose an efficient implementation of the IRLS method for a hybrid $l^1/l^2$ minimization problem that behaves as $l^2$ norm fit for small residual and $l^1$ norm fit for large residuals. The proposed algorithm shows more robust characteristics to the decision of the threshold value than the l1 norm IRLS inversion does with respect to the threshold value to avoid singularity.

• Journal of the Mineralogical Society of Korea
• /
• v.16 no.1
• /
• pp.65-73
• /
• 2003

In this study we have derived the two correction methods of hydrogen bonding distance. In case of the intermediate or long hydrogen bond(＞2.5 $\AA$), hydrogen bonding distances can be corrected by using the function d(O-H)=exp((2.173-d(O…O))/0.138)+0.958 obtained by least- squares fit to the data from the neutron diffraction at low temperatures. The valence-least-squares method is effective for the distance correction of very short hydrogen bond(<2.5 $\AA$). The distance correction is necessary for the long intermolecular hydrogen bond obtained from X-ray diffraction analysis.

• Wind and Structures
• /
• v.19 no.4
• /
• pp.371-387
• /
• 2014

Probability plotting positions are popular and used as the basis for distribution fitting and for inspecting the quality of the fit because of its simplicity. The plotting positions that lead to excellent approximation to the mean of the order statistics should be used if the objective of the fitting is to estimate quantiles. Since the mean depends on the sample size and is not amenable for simple to use closed form solution, many plotting positions have been presented in the literature, including a new plotting position that is derived based on the weighted least-squares method. In this study, the accuracy of using the new plotting position to fit the Gumbel distribution for estimating quantiles is assessed. Also, plotting positions derived by fitting the mean of the order statistics for all ranks is proposed, and an approximation to the covariance of the order statistics for the Gumbel (and Weibull) variate is given. Relative bias and root-mean-square-error of the estimated quantiles by using the proposed plotting position are shown. The use of the proposed plotting position to estimate the quantiles of annual maximum wind speed is illustrated.

• Journal of the Korean Data and Information Science Society
• /
• v.17 no.2
• /
• pp.609-617
• /
• 2006

A new fuzzy c-regression algorithm for switching regression analysis is presented, which combines fuzzy c-means clustering and least squares support vector machine. This algorithm can detect outliers in switching regression models while yielding the simultaneous estimates of the associated parameters together with a fuzzy c-partitions of data. It can be employed for the model-free nonlinear regression which does not assume the underlying form of the regression function. We illustrate the new approach with some numerical examples that show how it can be used to fit switching regression models to almost all types of mixed data.

• ETRI Journal
• /
• v.34 no.4
• /
• pp.641-644
• /
• 2012

The problem of sinusoidal parameter estimation at two channels with common frequency in white Gaussian noise is addressed. By making use of the linear prediction property, an iterative linear least squares (LLS) algorithm for accurate frequency estimation is devised. The remaining parameters are then determined according to the LLS fit with the use of the frequency estimate. It is proven that the variance of the frequency estimate achieves Cram$\acute{e}$r-Rao lower bound at sufficiently small noise conditions.

• Journal of the Korean earth science society
• /
• v.38 no.5
• /
• pp.333-344
• /
• 2017

We present the environmental dependence of the luminosity-size relation of galaxies in the local universe (z < 0.01) along with their dependence on galaxy morphology represented by five broad types (E, dEs, S0, Sp, and Irr). The environmental parameters we consider are the local background density and the group/cluster membership together with the clustercenteric distance for the Virgo cluster galaxies. We derive the regression coefficient (${\beta}$), i.e., the slope of the line representing the least-squares fitting to the data and the Pearson correlation coefficient (c.c.) representing the goodness of the least-squares fit along with the confidence interval from bootstrap resampling. We find no significant dependence of the luminosity-size relation on galaxy morphology. However, there is a weak dependence of the luminosity-size relations on the environment of galaxies, in the sense that galaxies in the low density environment have shallower slopes than galaxies in the high density regions except for elliptical galaxies that show an opposite trend.