• Journal of the Korean Data and Information Science Society
• /
• v.14 no.4
• /
• pp.965-974
• /
• 2003

In this paper we propose an incremental classification of multi-class data set by LS-SVM. By encoding the output variable in the training data set appropriately, we obtain a new specific output vectors for the training data sets. Then, online LS-SVM is applied on each newly encoded output vectors. Proposed method will enable the computation cost to be reduced and the training to be performed incrementally. With the incremental formulation of an inverse matrix, the current information and new input data are used for building another new inverse matrix for the estimation of the optimal bias and lagrange multipliers. Computational difficulties of large scale matrix inversion can be avoided. Performance of proposed method are shown via numerical studies and compared with artificial neural network.

• Proceedings of the Korean Society of Near Infrared Spectroscopy Conference
• /
• 2001.06a
• /
• pp.4106-4106
• /
• 2001

A method to discriminate human gallstones by nea. infrared(NIR) spectrometry using a soft independent modeling of class analogy (SIMCA) has been studied. The fifty NIR spectra of gallstones in the wavenumber range from 4500 to 10,000 cm$\^$-1/ were measured. The forty samples were classified to three classes, cholesterol stone, calcium bilirubinate stone and calcium carbonate stone according to the contents of major components in each gallstone. The training set which contained objects of the different known class was constructed using forty NIR spectra and the test set was made with ten different gallstone spectra. The number of important principal components(PCs) to describe the class was determined by cross validation in order to improve the decision criterion of the SIMCA for the training set. The score plots of the class training set whose objects belong to the other classes were inspected. The critical distance of each class was computed using both the Euclidean distance and the Mahalanobis distance at a proper level of significance(${\alpha}$). Two methods were compared with respect to classification and their robustness towards the number of PCs selected to describe different classes.

• PNF and Movement
• /
• v.18 no.1
• /
• pp.133-141
• /
• 2020

Purpose: The purpose of this study was to conduct balance training through vision control to improve the balance, postural control, and balance confidence and to decrease the visual and sensory dependence of stroke patients. Methods: Twenty-eight chronic stroke patients volunteered to participate in the study. They were randomly assigned to the eyes-closed and the eyes-open training groups. Three times a week for four weeks each group performed an unstable-support session and a balance training session for thirty minutes per set. Their balance, postural control, and balance confidence were assessed using BIO Rescue (BR), the postural assessment scale for stroke (PASS), and the Korean activity-specific balance confidence scale (K-ABC), respectively. All data were analyzed using SPSS version 22.0. Statistical methods before and after working around the average value of each dataset were independent T-test. The significance level for statistical analyses was set at 0.05. Results: Comparison between the groups showed statistically significant effects on all variables before and after the intervention (p < 0.05). Conclusion: This study reflected that balance-training programs involving vision control improve the balance, postural control, and balance confidence of chronic stroke patients. Thus, stroke patients should undergo training programs that increase the use of their other senses with vision control in clinical practice.

• Proceedings of the ESK Conference
• /
• 1996.10a
• /
• pp.197-205
• /
• 1996

Unification of neural network with a hierarchical pattern recognition is presented for recognizing large set of objects. A two-step identification procedure is developed for pattern recognition: coarse and fine identification. The coarse identification is designed for finding a class of object while the fine identification procedure is to identify a specific object. During the training phase a course neural network is trained for clustering larger set of reference objects into a number of groups. For training a fine neural network, expert neural network is also trained to identify a specific object within a group. The presented idea can be interpreted as two step identification. Experimental results are given to verify the proposed methodology.

• Journal of the Korean Data and Information Science Society
• /
• v.15 no.2
• /
• pp.499-506
• /
• 2004

The Unscented Kalman Filter is known to outperform the Extended Kalman Filter for the nonlinear state estimation with a significance advantage that it does not require the computation of Jacobian but EKF has a competitive advantage to the UKF on the performance time. We compare both algorithms on training the artificial neural network. The validation data set is used to estimate parameters which are supposed to result in better fitting for the test data set. Experimental results are presented which indicate the performance of both algorithms.

• The Mathematical Education
• /
• v.42 no.4
• /
• pp.481-501
• /
• 2003

The main purpose of this work is to propose programs of algebra courses for the department of mathematics education of teacher training universities. Set Theory, Linear Algebra, Number Theory, Abstract Algebra I, Abstract Algebra II, and Philosophy of Mathematics for School Teachers are discussed in this article.

• Journal of the Korea Convergence Society
• /
• v.5 no.1
• /
• pp.35-41
• /
• 2014

An automatic speech recognition system is one of the popular research problems. There are many research groups working in this field for different language including Japanese. Japanese vowel recognition is one of important parts in the Japanese speech recognition system. The vowel classification system with the Mamdani fuzzy inference system was developed in this research. We tested our system on the blind test data set collected from one male native Japanese speaker and four male non-native Japanese speakers. All subjects in the blind test data set were not the same subjects in the training data set. We found out that the classification rate from the training data set is 95.0 %. In the speaker-independent experiments, the classification rate from the native speaker is around 70.0 %, whereas that from the non-native speakers is around 80.5 %.

• Proceedings of the Korean Statistical Society Conference
• /
• 2003.10a
• /
• pp.137-141
• /
• 2003

LS-SVM(least squares support vector machine) is a widely applicable and useful machine learning technique for classification and regression analysis. LS-SVM can be a good substitute for statistical method but computational difficulties are still remained to operate the inversion of matrix of huge data set. In modern information society, we can easily get huge data sets by on line or batch mode. For these kind of huge data sets, we suggest an on line pruning regression method by LS-SVM. With relatively small number of pruned support vectors, we can have almost same performance as regression with full data set.

• Journal of Information Processing Systems
• /
• v.13 no.5
• /
• pp.1345-1357
• /
• 2017

A prototype selection method chooses a small set of training points from a whole set of class data. As the data size increases, the selected prototypes play a significant role in covering class regions and learning a discriminate rule. This paper discusses the methods for selecting prototypes in a classification framework. We formulate a prototype selection problem into a set covering optimization problem in which the sets are composed with distance metric and predefined classes. The formulation of our problem makes us draw attention only to prototypes per class, not considering the other class points. A training point becomes a prototype by checking the number of neighbors and whether it is preselected. In this setting, we propose a greedy algorithm which chooses the most relevant points for preserving the class dominant regions. The proposed method is simple to implement, does not have parameters to adapt, and achieves better or comparable results on both artificial and real-world problems.

• Journal of The Korean Society of Integrative Medicine
• /
• v.2 no.1
• /
• pp.91-100
• /
• 2014

Purpose : The purpose of this study was to know the influence of the difference in shoulder slope from balance training. Method : Training was divided that the 10 people were a Trampoline group, 10 people were a Togu group, 8 people were a Balance Board group and 9 people were a Control group. Method of the training was that the Trampoline group was carried out on the Trampoline. The training was carried out totally 2 times of 8 set per a week and had a break time during 10 seconds after carried out 30 seconds per one set. Togu group was carried out totally 2 times of 5 minutes per a week and had a break time during 30 seconds after carried out 2 minutes per one set. Balance Board group was carried out totally 2 times of 5 minutes per a week and had a break time 30 seconds after carried out 2 minutes per one set. Data was analyzed by repeated measure of one way ANOVA and repeated measure ANOVA. Result : The shoulder of the slope difference decreased significantly after balance training. The Trampoline group decreased from $3.13{\pm}1.01$ to $2.37{\pm}1.11$, the Togu group decreased from $3.78{\pm}0.85$ to $3.78{\pm}0.85$, the Balance Board group decreased from $1.78{\pm}0.82$ to $1.65{\pm}0.59$ and the Control group decreased from $1.77{\pm}1.16$ to $1.61{\pm}0.62$. Conclusion : The effectiveness improved in the order Togu group, Trampoline group, Balance Board group and Control group from the result of the balance training about difference of slope shoulder.