• Communications for Statistical Applications and Methods
• /
• v.30 no.4
• /
• pp.369-388
• /
• 2023

In this paper, we develop the two-step procedure that detects and estimates the position of structural changes for multivariate nonstationary time series, either on mean parameters or second-order structures. We first investigate the presence of mean structural change by monitoring data through the aggregated cumulative sum (CUSUM) type statistic, a sequential procedure identifying the likely position of the change point on its trend. If no mean change point is detected, the proposed method proceeds to scan the second-order structural change by modeling the multivariate nonstationary time series with a multivariate locally stationary Wavelet process, allowing the time-localized auto-correlation and cross-dependence. Under this framework, the estimated dynamic spectral matrices derived from the local wavelet periodogram capture the time-evolving scale-specific auto- and cross-dependence features of data. We then monitor the change point from the lower-dimensional approximated space of the spectral matrices over time by applying the dynamic principal component analysis. Different from existing methods requiring prior information on the type of changes between mean and covariance structures as an input for the implementation, the proposed algorithm provides the output indicating the type of change and the estimated location of its occurrence. The performance of the proposed method is demonstrated in simulations and the analysis of two real finance datasets.

• Communications for Statistical Applications and Methods
• /
• v.16 no.3
• /
• pp.511-518
• /
• 2009

We consider the problem of testing for change and estimating the unknown change-point in a sequence of time-ordered observations from the binomial and Poisson distributions. Including the likelihood ratio test, Gombay and Horvath (1990) tests are studied and the proposed change-point estimator is derived from their test statistic. A power study of tests and a comparison study of change-point estimators are done via simulation.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2011.05a
• /
• pp.252-252
• /
• 2011

Rainfall-runoff models are calibrated and validated by using a same data set such as observations. The past climate change effects the present rainfall pattern and also will effect on the future. To predict rainfall-runoff more preciously we have to consider the climate change pattern in the past, present and the future time. Thus, in this study, the climate change represents changes in mean precipitation and standard deviation in different patterns. In some river basins, there is no enough length of data for the analysis. Therefore, we have to generate the synthetic data using proper distribution for calculation of precipitation based on the observed data. In this study, Kajiyama model is used to analyze the runoff in the dry and the wet period, separately. Mean and standard deviation are used for generating precipitation from the gamma distribution. Twenty hypothetical scenarios are considered to show the climate change conditions. The mean precipitation are changed by -20%, -10%, 0%, +10% and +20% for the data generation with keeping the standard deviation constant in the wet and the dry period respectively. Similarly, the standard deviations of precipitation are changed by -20%, -10%, 0%, +10% and +20% keeping the mean value of precipitation constant for the wet and the dry period sequentially. In the wet period, when the standard deviation value varies then the mean NSE ratio is more fluctuate rather than the dry period. On the other hand, the mean NSE ratio in some extent is more fluctuate in the wet period and sometimes in the dry period, if the mean value of precipitation varies while keeping the standard deviation constant.

• Journal of Korean Physical Therapy Science
• /
• v.8 no.1
• /
• pp.893-906
• /
• 2001

The purposes of this study were to compare pelvic tilt. range of motion(ROM) of hip rotation, and leg length difference before and after manipulation and to investigate correlation between changes of each variables after manipulation of sacroiliac pint in 31 low back pain patients(11 males, 20 females) with sacroiliac pint dysfunction. The sacroiliac pint of patients was manipulated on the side of anterior pelvic tilt, using the technique described by Stoddard(1962) and Greenmann (1996). I used this technique because it usually eliminated sacroiliac Pint dysfunction in one treatment session. SPSS for window computer system was used to analyze the data. Also t-test was performed for comparison of the variables before and after manipulation, and Pearson product-moment correlation analysis and regression analysis were performed for changes of each variables after manipulation. The result were as follows: 1. The pelvic tilt after manipulation was significantly decreased(mean=$2.79^{\circ}$) compared with the pelvic tilt before manipulation(p=.001). 2. The PROM of hip internal rotation ipsilateral to anterior pelvic tilt after manipulation significantly decreased (mean = $1.88^{\circ}$) compared with hip internal rotation before manipulation (p=.008). The PROM of hip internal rotation ipsilateral to posterior pelvic tilt after manipulation significantly increased(mean = $1.29^{\circ}$) compared with hip internal rotation before manipulation (p=.029). 3. The PROM of hip external rotation ipsilateral to anterior pelvic tilt after manipulation significantly increased(mean=$2.42^{\circ}$) compared with the hip external rotation before manipulation(p=$2.42^{\circ}$) compared with the hip external rotation ipsilateral to posterior pelvic tilt after manipulation significantly decreased(mean = $1.84^{\circ}$) compared with the hip external rotation before manipulation (p=.008). 4. Leg length difference after manipulation significantly decreased(mean=2.15 mm) compared with leg length difference before manipulation (p=.008). Regression analysis revealed that a fair correlation was found between change in leg length difference and change in anterior pelvic tilt after manipulation(p=.009). 5. Pearson product-moment correlation coefficient was used to assess differences of the variables after manipulation. A fair correlation was found between change in leg length difference and change in anterior pelvic tilt after manipulation(r=.462, p<.01). A fair correlation was found between change in anterior pelvic tilt and change in hip internal rotation ipsilateral to anterior pelvic tilt(r=.397, p<.05) and between change in anterior pelvic tilt and change in hip external rotation ipsilateral to anterior pelvic tilt(r=.516, p<.01). A fair correlation was found between change in posterior pelvic tilt and changes in hip internal rotation ipsilateral to posterior pelvic tilt (r=.441, p<.05) and between change in posterior pelvic tilt and change in hip external rotation ipsilateral to posterior pelvic tilt(r=.361, p<.05). A fair correlation was found between change in hip internal rotation ipsilateral to anterior pelvic tilt and change in hip external rotation ipsilateral to posterior pelvic tilt(r=.388, p<.05) and between change in hip internal rotation ipsilateral to posterior pelvic tilt and change in hip internal rotation ipsilateral to anterior pelvic tilt(r=.426. p<.05).

• The Journal of Information Systems
• /
• v.14 no.3
• /
• pp.117-122
• /
• 2005

This paper presents a method for detecting scene changes in video sequences, in which the $chi^{2}$-test is modified by imposing weights according to NTSC standard. To automatically determine threshold values for scene change detection, the proposed method utilizes the frame differences that are obtained by the weighted $chi^{2}$-test. In the first step, the mean and the standard deviation of the difference values are calculated, and then, we subtract the mean difference value from each difference value. In the next step, the same process is performed on the remained difference values, mean-subtracted frame differences, until the stopping criterion is satisfied. Finally, the threshold value for scene change detection is determined by the proposed automatic threshold estimation algorithm. The proposed method is tested on various video sources and, in the experimental results, it is shown that the proposed method is reliably estimates the thresholds and detects scene changes.

• Korean Journal of Community Nutrition
• /
• v.11 no.3
• /
• pp.346-360
• /
• 2006

This study was conducted to evaluate the effects of home-visiting nutrition education for the elderly with high fasting glucose level in an urban community. The study subjects were 40 elderly people, whose information on general characteristics, anthropometric measurement, biochemical indices, nutrition knowledge, nutrition attitude, dietary habit, food intake and nutrient intake were obtained at baseline. The education group received 6 weekly visits of home-visiting nutrition education from 15 March to 25 April 2004. In the baseline-survey, the education group and non-education group showed no differences in their general characteristics, health-related characteristics, anthropometric measurements, biochemical indices, nutrition knowledge, nutrition attitudes, dietary habits, and food and nutrient intake levels. The difference of mean change of nutrition knowledge, nutrition attitude and dietary habit after home-visiting nutrition education had been studied. The nutrition knowledge score increased by 1.4 in the education group; however, those in the non-education group which increased by 0.4, thus, the differences of mean change were statistically significant (p < 0.05). The nutrition attitude score increased by 1.2 in the education group; however, those in the non-education group decreased by 0.4, thus, the differences of mean change were statistically significant (p < 0.01). The dietary habit score increased by 1.7 in the education group; however, those in the non-education group decreased by 2.8, thus, the differences of mean change were statistically significant (p < 0.01). The difference of mean change of anthropometric indices and biochemical indices in education group and the non-education group was not significant. Looking over the zcereals and their products, vegetables, seaweeds, meats and their products, and fish than the non-education group. The MAR increased by 0.06 in the education group; however, that in the non-education group increased by 0.01, thus, the differences of mean change were statistically significant (p < 0.05). The differences of mean change of fasting blood glucose and biochemical indices after home-visiting nutrition education were studied. Fasting blood glucose decreased by 7.6 mg/dL in the education group; however, in the non-education group which increased by 0.4 mg/dL, the difference of mean change was not significant (p = 0.051). The above findings suggest that home visiting nutrition education increases the nutrition knowledge and nutrition attitude, as well as, it is effective to change dietary habits. If the education period is extended, not fasting blood glucose improvement was insignificant, but fasting blood glucose improvement ability could be found by changing dietary habits.

• Journal of Korean Society for Quality Management
• /
• v.26 no.4
• /
• pp.101-110
• /
• 1998

The mean residual life function is the expected remaining life of an item at age $\chi$. The problem of trend change in the mean residual life is great interest in the reliability and survival analysis. In this paper we develop a new test statistic for testing whether or not the mean residual life changes its trend based on a complete sample. Monte Carlo simulations are conducted to compare the perfor mance of our test statistic with that of previously known tests.

• Journal of Korean Society for Quality Management
• /
• v.30 no.3
• /
• pp.195-202
• /
• 2002

The problem of trend change in the mean residual life is great Interest in the reliability and survival analysis. In this paper, a new test statistic for testing whether or not the mean residual life changes its trend Is developed. It is assumed that neither the change point nor the proportion at which the trend change occurs is known. The asymptotic null distribution of test statistic is established and asymptotic critical values of the asymptotic null distribution is obtained. Monte Carlo simulation is used to compare the proposed test with previously known tests.

• Journal of Korean Society for Quality Management
• /
• v.30 no.4
• /
• pp.78-85
• /
• 2002

The problem of trend change in the mean residual life is great interest in the reliability and survival analysis. In this paper, a new test statistic for testing whether or not the mean residual life changes its trend is developed. It is assumed that neither the change point nor the proportion at which the trend change occurs is known. The asymptotic null distribution of test statistic is established and asymptotic critical values of the asymptotic null distribution is obtained. Monte Carlo simulation is used to compare the proposed test with previously known tests.

• Journal of radiological science and technology
• /
• v.40 no.3
• /
• pp.415-421
• /
• 2017

In the case of radiation therapy for prostate cancer, a balloon infused with a certain amount of air through the anus is used to reduce rectal dose. Because of the reason, radiation therapy for prostate cancer has acquired CBCT for daily image induction. In order to maintain the anatomical structure most similar to the first CT taken before treatment, it is pretreated, but it can not be said to be perfectly consistent. In two actual treatment regimens, the volume of the bladder was measured as 45.82 cc and 63.43 cc, and the equivalent diameter was 4.4 cm and 4.9 cm. As a result of this study, the mean volume of the bladder was estimated to be 56.2 cc, 105.6 cc by 20 CBCT. The mean dose of CBCT was 1.74% and the mean Bladder mean dose was 96.67%. In case B, PTV mean dose was 4.31%, Bladder mean Dose was estimated to be 97.35%. The changes in the volume of the bladder resulted in changes in the dose of PTV and bladder. The correlation coefficient of bladder dose according to the change of bladder volume showed linearity of mean dose $R^2=-0.94$. The correlation coefficient of the PTV dose according to the volume change of the bladder showed linearity of mean dose $R^2=0.04$. It was found that the dose change of PTV was larger than that of bladder according to the change of bladder volume.