• The Journal of the Acoustical Society of Korea
• /
• v.29 no.6
• /
• pp.374-380
• /
• 2010

It is known that the problem of FIR filtering with noisy input and output data can be solved by a total least squares (TLS) estimation. It is also known that the performance of the TLS estimation is very sensitive to the ratio between the variances of the input and output noises. In this paper, we propose a convex combination algorithm between the ordinary recursive LS based TLS (RTLS) and the ordinary recursive LS (RLS). This combined algorithm is robust to the noise variance ratio and has almost the same complexity as the RTLS. Simulation results show that the proposed algorithm performs near TLS in noise variance ratio ${\gamma}{\approx}1$ and that it outperforms TLS and LS in the rage of 2 < $\gamma$ < 20. Consequently, the practical workability of the TLS method applied to noisy data has been significantly broadened.

• Communications for Statistical Applications and Methods
• /
• v.13 no.2
• /
• pp.233-241
• /
• 2006

Classical principal component analysis (PCA) can be formulated as finding the linear subspace that best accommodates multidimensional data points in the sense that the sum of squared residual distances is minimized. As alternatives to such LS (least squares) fitting approach, we produce LMS (least median of squares) and LTS (least trimmed squares)-type PCA by minimizing the median of squared residual distances and the trimmed sum of squares, in a similar fashion to Rousseeuw (1984)'s alternative approaches to LS linear regression. Proposed methods adopt the data-driven optimization algorithm of Croux and Ruiz-Gazen (1996, 2005) that is conceptually simple and computationally practical. Numerical examples are given.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2003.10a
• /
• pp.15-19
• /
• 2003

The Total Least Squares (TLS) method is introduced in comparison with the conventional Least Squares (LS) method. The principles and mathematical models for both methods are summarized and the comparison results from their applications to a simple geometric example, fitting a straight line to a set of 2D points are presented. As conceptually reasoned, the results clearly indicate that LS is more susceptible of producing wrong parameters with worse precision rather than TLS. For many applications in surveying, can adjustment computation and parameter estimation based on TLS provide better results.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.2
• /
• pp.363-369
• /
• 2010

In this paper we propose a mixed-effects least squares support vector regression (LS-SVR) for longitudinal data. We add a random-effect term in the optimization function of LS-SVR to take random effects into LS-SVR for analyzing longitudinal data. We also present the model selection method that employs generalized cross validation function for choosing the hyper-parameters which affect the performance of the mixed-effects LS-SVR. A simulated example is provided to indicate the usefulness of mixed-effect method for analyzing longitudinal data.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.13 no.4
• /
• pp.486-490
• /
• 2003

Software qualify models can predict the risk of faults in the software early enough for cost-effective prevention of problems. This paper introduces a least squares support vector machine (LS-SVM) as a fuzzy regression method for predicting fault ranges in the software under development. This LS-SVM deals with the fuzzy data with crisp inputs and fuzzy output. Predicting the exact number of bugs in software is often not necessary. This LS-SVM can predict the interval that the number of faults of the program at each session falls into with a certain possibility. A case study on software reliability problem is used to illustrate the usefulness of this LS -SVM.

• Proceedings of the IEEK Conference
• /
• 2000.06a
• /
• pp.45-48
• /
• 2000

The use of multi-level modulation scheme in the wireless LAN(Local Area Networks) system requires an accurate channel estimation. In this paper, we present sequential least squares(LS) channel estimation scheme based on decision-directed channel tracking scheme. The proposed scheme improves the performance of the conventional LS estimator for wireless LAN. In addition, its structure is suitable for the high-rate wireless LAN. Simulation results show that the proposed scheme achieves about IdB Packet Error Rate(PER) gain compared to the LS scheme in a frequency selective channel.

• Journal of the Korean Data and Information Science Society
• /
• v.15 no.2
• /
• pp.507-513
• /
• 2004

Least-squares support vector machine (LS-SVM) has been very successful in pattern recognition and function estimation problems for crisp data. In this paper, we propose LS-SVM approach to evaluating fuzzy regression model with multiple crisp inputs and a Gaussian fuzzy output. The proposed algorithm here is model-free method in the sense that we do not need assume the underlying model function. Experimental result is then presented which indicate the performance of this algorithm.

• Journal of the Korean Data and Information Science Society
• /
• v.27 no.3
• /
• pp.827-833
• /
• 2016

In this paper, we propose a deep least squares support vector machine (LS-SVM) for regression problems, which consists of the input layer and the hidden layer. In the hidden layer, LS-SVMs are trained with the original input variables and the perturbed responses. For the final output, the main LS-SVM is trained with the outputs from LS-SVMs of the hidden layer as input variables and the original responses. In contrast to the multilayer neural network (MNN), LS-SVMs in the deep LS-SVM are trained to minimize the penalized objective function. Thus, the learning dynamics of the deep LS-SVM are entirely different from MNN in which all weights and biases are trained to minimize one final error function. When compared to MNN approaches, the deep LS-SVM does not make use of any combination weights, but trains all LS-SVMs in the architecture. Experimental results from real datasets illustrate that the deep LS-SVM significantly outperforms state of the art machine learning methods on regression problems.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.5
• /
• pp.959-965
• /
• 2010

For multiclassification, it is often the case that some variables are not important while some variables are more important than others. We propose a novel algorithm for selecting such relevant variables for multiclassification. This algorithm is base on multiclass least squares support vector machine (LS-SVM), which uses results of multiclass LS-SVM using one-vs-all method. Experimental results are then presented which indicate the performance of the proposed method.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.2C
• /
• pp.255-261
• /
• 2004

The problem of geolocation using multiple satellites is to determine the position of a transmitter located on the Earth by processing received signals. The specific problem addressed in this paper is that of estimating the position of a stationary transmitter located on or above the Earth's surface from measured time difference of arrivals (TDOA) by a geostationary orbiting (GSO) satellite and a low earth orbiting (LEO) satellite. The proposed geolocation method is based on the total least squares (TLS) algorithm. Under erroneous positions of the satellites together with noisy TDOA measurements, the TLS algorithm provides a better solution. By running Monte-Carlo simulations, the proposed method is compared with the ordinary least squares (LS) approach.