• Journal of the Korean Data and Information Science Society
• /
• v.14 no.1
• /
• pp.67-74
• /
• 2003

Median regression analysis has robustness properties which make it an attractive alternative to regression based on the mean. Support vector machine (SVM) is used widely in real-world regression tasks. In this paper, we propose a new SV median regression based on check function. And we illustrate how this proposed SVM performs and compare this with the SVM based on absolute deviation loss function.

• Communications for Statistical Applications and Methods
• /
• v.30 no.2
• /
• pp.135-147
• /
• 2023

Skipped median is more robust than median when outliers are not symmetrically distributed. In this work, we propose a novel algorithm to estimate the skipped median. The idea of skipped median and the new algorithm are extended to regression problem, which is called least clipped absolute deviation (LCAD). Since our proposed algorithm for nonconvex LCAD optimization makes use of convex least absolute deviation (LAD) procedure as a subroutine, regularizations developed for LAD can be directly applied, without modification, to LCAD as well. Numerical studies demonstrate that skipped median and LCAD are useful and outperform their counterparts, median and LAD, when outliers intervene asymmetrically. Some extensions of the idea for skipped median and LCAD are discussed.

• Physical Therapy Korea
• /
• v.10 no.3
• /
• pp.141-149
• /
• 2003

Median frequency can be regarded as a valid indicator of local muscle fatigue. As local muscle fatigue develops, the muscle fiber conduction velocity decreases, the fast twitch fibers are recruited less, and consequently the median frequency shifts toward the lower frequency area. The aim of this study was to test the characteristics of the median frequency according to exercise load (30% and 60% of MVC on the biceps brachii, 40% and 80% of MVC on the vastus lateralis) during the fatiguing isometric exercise. Thirteen healthy male volunteer students of Yonsei University were recruited. After the testing maximal voluntary isometric contraction, three variables (initial median frequency, regression slope, fatigue index) from the regression line of MDF data were measured in each exercise load. The results showed that the regression slope and fatigue index were significantly different for the biceps brachii, but not for the vastus lateralis initial MDF was not significant difference according to the exercise load on both muscles. The regression slope and fatigue index could monitor physiologic muscle change during fatiguing isometric exercise. The results showed that two MDF variables reflect the local muscle fatigue according to the exercise load.

• Journal of applied mathematics & informatics
• /
• v.27 no.3_4
• /
• pp.903-908
• /
• 2009

The shrink parameter in ridge regression may be contaminated by outlying points. We propose robust cross validation scores in ridge regression instead of classical cross validation. We use robust location estimators such as median, least trimmed squares, absolute mean for robust cross validation scores. The robust scores have global robustness. Simulations are performed to show the effectiveness of the proposed estimators.

• Journal of Korea Multimedia Society
• /
• v.20 no.10
• /
• pp.1671-1677
• /
• 2017

In this paper, we propose a machine learning method for random noise removal in image deblurring. The proposed method uses a logistic regression to select reliable data to use them, and, at the same time, to exclude data, which seem to be corrupted by random noise, in the deblurring process. The proposed method uses commonly available images as training data. Simulation results show an improved performance of the proposed method, as compared with the median filtering based reliable data selection method.

• Journal of Biomedical Engineering Research
• /
• v.24 no.3
• /
• pp.175-181
• /
• 2003

Studies of muscle fatigue have mostly been checked under isometric exercise. However EMG signals from isotonic exercise generate uncontrollable noise, so there were difficulties in gathering reliable median frequency and muscle fatigue index if frequency analysis was equally applied in isometric exercise. This study tried to compare the differences of muscle fatigue determinant variables in terms of median frequency searching methods of EMG signal, which was estimated in isotonic exercise. To accomplish this, we determined median frequency by using different FFT intervals and overlapping ration of consecutive FFT sections under the same EMG signal, and then searched for a linear regression line, and compared initial median frequency, slope, and muscle fatigue index which were variables under the linear regression line. In result of comparison, initial median frequency was more elevated as FFT exercise interval became larger. The slope of the linear regression line showed distinguishable decreasing tendency as FFT intervals were larger and overlapping sections were smaller. Significant tendency of muscle fatigue index in FFT interval was shown by high muscle fatigue index in specific FFT intervals.

• Journal of the Korean Statistical Society
• /
• v.10
• /
• pp.105-121
• /
• 1981

The problem of selection of a subset containing the largest of several slope parameters of regression equations is considered. The proposed selection procedure is based on the weighted median estimators for regression parameters and the median of rescaled absolute residuals for scale parameters. Those estimators are compared with the classical least squares estimators by a simulation study. A Monte Carlo comparison is also made between the new procedure based on the weighted median estiamtors and the procedure based on the least squares estimators. The results show that the proposed procedure is quite robust with respect to the heaviness of distribution tails.

• Physical Therapy Korea
• /
• v.8 no.2
• /
• pp.41-54
• /
• 2001

Fatigue is the decline in force produced as a result of prolonged muscle activity. Localized muscle fatigue can be identified by a shift toward low in the frequency components of the EMG signal, typically represented by a fall in the median frequency. Previous studies show that a shortened muscle develops a higher fatigue than elongated muscles. The purpose of this study was to investigate the time-related change of median frequency and torque during maximal isometric back extension exercises at different exercise angles ($0^{\circ}$, $12^{\circ}$, $36^{\circ}$, $72^{\circ}$). Twenty healthy subjects (mean age = $24.35{\pm}2.70$) were evaluated in this study. Median frequency was extracted from the EMG signals by fast Fourier transform (FFT). Initial median frequency and the slope of median frequency change over time were computed from linear regression analysis. Pearson's product moment correlation was used to quantify the relationship between slope of median frequency and torque. The results were as follows: 1) Significant differences in y-intercepts of torque regression equation with respect to exercise angle were shown. However, there were no differences in the slopes of the median frequency and torque, and y intercept of the median frequency among exercise angles. 2) There was no significant correlation between slope of median frequency and torque. 3) But there was moderate correlation between median frequency and torque at each exercise angle. In conclusion, the exercise angle during maximal isometric back extension exercise is not a direct effect on slope of median frequency and torque. But results showed that median frequency and torque shift were highly correlated in all subjects.

• Physical Therapy Korea
• /
• v.8 no.3
• /
• pp.63-76
• /
• 2001

Many studies have shown that the initial median frequency (MDF) and slope correlate with the muscle fiber composition. This study tested the hypothesis that the initial MDF and slope are fixed, regardless of the interval at which data are collected. MDF data using moving fast Fourier transformation of EMG signals, following local fatigue induced by isotonic exercise, were obtained. An inverse FFT was used to eliminate noise, and characteristic decreasing regression lines were obtained. The regression analysis was done in three different periods, the first one third, first half, and full period, looking at variance in the initial MDF, slope, and fatigue index. Data from surface EMG signals during fatiguing isotonic exercise of the biceps brachii and vastus lateralis in 20 normal subjects were collected. The loads tested were 30% and 60% maximum voluntary contraction (MVC) in the biceps brachii and 40% and 80% MVC in the vastus lateralis. The rate was 25 flexions per minute. There were no significant differences in the initial MDF or slope during the early or full periods of the regression, but there was a significant difference in the fatigue index. Therefore, to observe the change in the initial MDF and slope of the MDF regression line during isotonic exercise, this study suggest that only the early interval need to be observed.

• Journal of Preventive Medicine and Public Health
• /
• v.41 no.6
• /
• pp.434-441
• /
• 2008

Objectives: This study was conducted to measure the decline in the health-related quality of life (HRQoL) associated with some diseases in South Korean adults. Methods: The EQ-5D health states in the 2005 National Health and Nutrition Examination Survey (NHNES) and the Korean EQ-5D valuation set were used to obtain the EQ-5D indexes of the study subjects. Each disease group was defined when the subjects reported to the NHNES that they were diagnosed with the corresponding disease during the previous 1 year by physicians. Since the distributions of the EQ-5D indexes in each subgroup were negatively skewed, median regression analysis was used to estimate the effects of specific diseases on the HRQoL. Median regression analysis produced estimates that approximated the median of the EQ-5D indexes and there are more robust for analyzing data with many outliers. Results: A total of 16,692 subjects (6,667 patients and 10,025 people without any disease) were included in the analysis. As a result of the median regression analysis, stroke had the strongest impact on the HRQoL for both males and females, followed by osteoporosis, osteoarthritis, rheumatic arthritis, and herniation of an intervertebral disc. While asthma had a significant impact on the HRQoL only in men, cataract, temporo-mandibular dysfunction, and peptic ulcer significantly affected the HRQoL only in women. Conclusions: Stroke and musculoskeletal diseases were associated with the largest losses of the HRQoL in Korean adults.