• IE interfaces
• /
• v.17 no.1
• /
• pp.1-12
• /
• 2004

In a modern semiconductor device manufacturing industry, statistical bin limits on wafer level test bin data are used for minimizing value added to defective product as well as protecting end customers from potential quality and reliability excursion. Most wafer level test bin data show skewed distributions. By Monte Carlo simulation, this paper evaluates methods and sample size effect regarding determination of statistical bin limits. In the simulation, it is assumed that wafer level test bin data follow the Poisson distribution. Hence, typical shapes of the data distribution can be specified in terms of the distribution's parameter. This study examines three different methods; 1) percentile based methodology; 2) data transformation; and 3) Poisson model fitting. The mean square error is adopted as a performance measure for each simulation scenario. Then, a case study is presented. Results show that the percentile and transformation based methods give more stable statistical bin limits associated with the real dataset. However, with highly skewed distributions, the transformation based method should be used with caution in determining statistical bin limits. When the data are well fitted to a certain probability distribution, the model fitting approach can be used in the determination. As for the sample size effect, the mean square error seems to reduce exponentially according to the sample size.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.6
• /
• pp.1553-1567
• /
• 1996

The numerous applications surface interpolation include the modeling and visualization phenomena. A tetrahedrization is one of pre-processing steps for 4-D space interpolation. The quality of a piecewise linear interpolation 4-D space depends not only on the distribution of the data points in $R^2$, but also on the data values. We show that the quality of approximation can be improved by data dependent tetraheadrization through visualization of 4-D space. This paper discusses Delaunary tetrahedrization method(sphere criterion) and one of the data dependent tetrahedrization methods(least squares fitting criterion). This paper also discusses new data dependent criteria:1) gradient difference, and 2) jump in normal direction derivative.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.26 no.10
• /
• pp.1961-1970
• /
• 2002

A laser scanner is widely used fur a device fur acquiring point data in reverse engineering. It is more efficient to generate a surface automatically from the line-typed data than scattered data of points clouds. In the case of a compound model, it is hard to represent all the scanned data into one surface maintaining its original line characteristics. In this paper, a method is presented to generate a surface by the segmentation of measured point data. After forming triangular net, the segmentation is done by the user input such as the angle between triangles, the number of facets to be considered as small segment, and the angle for combining small segment. B-spline fitting is implemented to the point data in each segment. The surface generation through segmentation shows a reliable result when it is applied to the models with curvature deviation regions. An useful algorithm for surface reconstruction is developed and verified by applying an practical model and shows a good tools fur reverse engineering in design modification.

• Journal of the Korean Statistical Society
• /
• v.5 no.1
• /
• pp.19-28
• /
• 1976

In empirical study for fitting a multiple linear regression model for successive cross-sections data observed on the same set of independent variables over several time periods, one often faces the problem of poor $R^2$, the multiple coefficient of determination, which provides a standard measure of how good a specified regression line fits the sample data.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.49 no.9
• /
• pp.362-369
• /
• 2012

Recently, a target detection based on a distributed sensor network has been much studied in active sonar. Zhou et al. proposed a target localization method using line fitting based on a distributed sensor network which consists of low complexity sensors that only report binary detection results. This method has three advantages relative to ML estimator. First, there is no need to estimate propagation model parameters. Second, the computation is simple. Third, it only use sensors with "detection", which implies less data to be collected by data processing center. However, this method has larger target localization error than the ML estimator. In this paper, a target localization method which modifies Zhou's method is proposed for reducing the localization error. The modified method shows the performance improvement that the target localization error is reduced by 40.7% to Zhou's method in the point of RMSE.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.4D
• /
• pp.555-561
• /
• 2009

In lieu of section profile data, a fitting of the bored tunnel shape is more significant confirmation for maintenance of a tunnel. Before the permit on the completion of a tunnel, deformation of the completed tunnel with respect to the design model are considered. And deformation can be produced at continuously along the entire of the tunnel section. This study firstly includes an analysis of algebraic approach and test it with an observed field data. And then a number of methods, line search method, genetic algorithm, and pattern search methods, are compared with the 3D tunnel shape fitting. Algebraic methods can solve a simple circular cylinder type as like a railway tunnel. However, a more complex model (compound circular curve and non circular) as like a highway tunnel has to be solved with soft computing tools in the cause of conditional constraints. The genetic algorithm and pattern search methods are computationally more intensive, but they are more flexible at a complex condition. The line search method is fastest, but it needs a narrow bounds of the initial values.

• Journal of Digital Convergence
• /
• v.19 no.10
• /
• pp.253-263
• /
• 2021

Machine learning is the process of constructing a cost function using learning data used for learning and an artificial neural network to predict the data, and finding parameters that minimize the cost function. Parameters are changed by using the gradient-based method of the cost function. The more complex the digital signal and the more complex the problem to be learned, the more complex and deeper the structure of the artificial neural network. Such a complex and deep neural network structure can cause over-fitting problems. In order to avoid over-fitting, a weight decay regularization method of parameters is used. We additionally use the value of the cost function in this method. In this way, the accuracy of machine learning is improved, and the superiority is confirmed through numerical experiments. These results derive accurate values for a wide range of artificial intelligence data through machine learning.

• Geomechanics and Engineering
• /
• v.36 no.6
• /
• pp.601-618
• /
• 2024

The joint probability distribution of uncertain geomechanical parameters of geotechnical strata is a crucial aspect in constructing the reliability functional function for roof structures. However, due to the limited number of on-site exploration and test data samples, it is challenging to conduct a scientifically reliable analysis of roof geotechnical strata. This study proposes a Copula method based on small sample exploration and test data to construct the intensity characteristics of roof geotechnical strata. Firstly, the theory of multidimensional copula is systematically introduced, especially the construction of four-dimensional Gaussian copula. Secondly, data from measurements of 176 groups of geomechanical parameters of roof geotechnical strata in 31 coal mines in China are collected. The goodness of fit and simulation error of the four-dimensional Gaussian Copula constructed using the Pearson method, Kendall method, and Spearman methods are analyzed. Finally, the fitting effects of positive and negative correlation coefficients under different copula functions are discussed respectively. The results demonstrate that the established multidimensional Gaussian Copula joint distribution model can scientifically represent the uncertainty of geomechanical parameters in roof geotechnical strata. It provides an important theoretical basis for the study of reliability functional functions for roof structures. Different construction methods for multidimensional Gaussian Copula yield varying simulation effects. The Kendall method exhibits the best fit in constructing correlations of geotechnical parameters. For the bivariate Copula fitting ability of uncertain parameters in roof geotechnical strata, when the correlation is strong, Gaussian Copula demonstrates the best fit, and other Copula functions also show remarkable fitting ability in the region of fixed correlation parameters. The research results can offer valuable reference for the stability analysis of roof geotechnical engineering.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.36 no.3
• /
• pp.269-281
• /
• 2012

This study investigates the influence of technology readiness, fashion innovativeness, and participation level perception on the acceptance intention of 3D virtual fitting systems. We presented a 3D virtual fitting system with detailed information that was watched by respondents who subsequently completed a research questionnaire. The data were collected from 300 subjects with an age range of 21 to 39 who have experienced Internet shopping. Descriptive statistics, Cronbach's alpha, factor analysis, correlation analysis, and multiple regression analysis were conducted. The results were as follows. First, fashion innovativeness, technology innovativeness, participation level perception, and optimism significantly influenced the acceptance intention. Second, fashion innovativeness, technology innovativeness, participation level perception, and optimism positively influenced the acceptance intention in the male group; however, technology innovativeness, participation level perception, optimism, and insecurity significantly influenced the acceptance intention in the female group. The results indicated that a marketing strategy has to be designed that focuses on consumers with high technology, fashion innovativeness, and optimism to increase the acceptance intention. In addition, markers have to enhance a participation level perception that will contribute to the introduction of 3D virtual fitting systems. Another notable finding was the importance to differentiate marketing strategies according to gender.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.36 no.11
• /
• pp.1174-1185
• /
• 2012

This study develops tight-fitting upper clothing to measure electrocardiography (ECG) data. Taking into consideration the elasticity of the clothing, we made 4 experimental clothes by applying to each a weft reduction rate of 40%, 50%, 60%, and 70%. The 4 experimental clothes were used to measure resting ECG, exercise ECG, and post-exercise ECG for 4 men in their 20s. We compared clothing pressures using sensors on the human body and on a dressform. Subjective wear sensations of the 4 experimental clothes were evaluated using a subjective 7-point scale (with 7 being most excellent). We measured clothing pressures by using the air type pressure (AMI 3037-2) for upper and lower chest sensors in the developed tight-fitting upper clothing. The lower chest sensor showed that the clothing pressure on a human body and dressform changed consistently as the weft reduction rate decreased. The upper chest sensor showed inconsistent changes in clothing pressure as the weft reduction rate decreased. The wearing-test result for preliminary subjects showed that the lower chest sensor was more stable than the upper chest sensor; therefore, we inserted the sensor at the lower chest position before performing ECG. Except for Subject 4, the resting ECGs were stably measured for 3 subjects (Subject 1, Subject 2, and Subject 3) in all the developed clothes (A clothing, B clothing, C clothing, and D clothing). However, D clothing showed stable ECG values after exercise. The results of the experiment showed that we could measure ECG without difficulty using clothes with a weft reduction rate of 40% when the movement was not intense; however, tight-fitting upper clothing with a weft reduction rate of 70% was necessary to measure exercise ECG and post-exercise ECG values.