• Communications for Statistical Applications and Methods
• /
• v.7 no.3
• /
• pp.817-827
• /
• 2000

In this paper, we investigate the Bayesian approach to random effect binomial regression models with improper prior due to the absence of information on parameter. We also propose a method of estimating the posterior moments and prediction and discuss some general methods for studying model assessment. The methodology is illustrated with Crowder's Seeds Data. Markov Chain Monte Carlo techniques are used to overcome the computational difficulties.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.11a
• /
• pp.244-247
• /
• 2020

Data augmentation has been helpful in improving the performance in deep learning, when we have a limited data and random erasing is one of the augmentations that have shown impressive performance in deep learning in multiple domains. But the main issue is that sometime it loses good features when randomly selected region is erased by some random values, that does not improve performance as it should. We target that problem in way that good features should not be lost and also want random erasing at the same time. For that purpose, we introduce new augmentation technique named Intra-Class Random Erasing (ICRE) that focuses on data to learn robust features of the same class samples by randomly exchanging randomly selected region. We perform multiple experiments by using different models including resnet18, VGG16 over variety of the datasets including ESC10, UrbanSound8K. Our approach has shown effectiveness over others methods including random erasing.

• Computers and Concrete
• /
• v.33 no.2
• /
• pp.147-162
• /
• 2024

This study presents a random-aggregate mesoscale model integrating the random distribution of the coarse aggerates and the damage mechanics of the mortar and interfacial transition zone (ITZ). This mesoscale model can generate the random distribution of the coarse aggregates according to the prescribed particle size distribution which enables the automation of the current methodology with different coarse aggregates' distribution. The main innovation of this work is to propose the "correction factor" to eliminate the dimensionally dependent mesh sensitivity of the concrete damaged plasticity (CDP) model. After implementing the correction factor through the user-defined subroutine in the randomly meshed mesoscale model, the predicted fracture resistance is in good agreement with the average experimental results of a series of geometrically similar single-edge-notched beams (SENB) concrete specimens. The simulated cracking pattern is also more realistic than the conventional concrete material models. The proposed random-aggregate mesoscale model hence demonstrates its validity in the application of concrete fracture failure and statistical size effect analysis.

• Communications for Statistical Applications and Methods
• /
• v.21 no.5
• /
• pp.423-433
• /
• 2014

Test statistics using cumulative sums of residuals have been widely used in various regression models including generalized linear models(GLM). Recently, Pan and Lin (2005) extended this testing procedure to the generalized linear mixed models(GLMM) having random effects, in which we encounter difficulties in computing the marginal likelihood that is expressed as an integral of random effects distribution. The Gaussian quadrature algorithm is commonly used to approximate the marginal likelihood. Many commercial statistical packages provide an option to apply this type of goodness-of-fit test in GLMs but available programs are very rare for GLMMs. We suggest a computational algorithm to implement the testing procedure in GLMMs by a freely accessible R package, and also illustrate through practical examples.

• Structural Engineering and Mechanics
• /
• v.6 no.5
• /
• pp.473-484
• /
• 1998

The first and second moments of response variables for SDOF systems with hysteretic nonlinearity are obtained by a direct linearization procedure. This adaptation in the implementation of well-known statistical linearization methods, provides concise, model-independent linearization coefficients that are well-suited for numerical solution. The method may be applied to systems which incorporate any hysteresis model governed by a differential constitutive equation, and may be used for zero or non-zero mean random vibration. The implementation eliminates the effort of analytically deriving specific linearization coefficients for new hysteresis models. In doing so, the procedure of stochastic analysis is made independent from the task of physical modeling of hysteretic systems. In this study, systems with three different hysteresis models are analyzed under various zero and non-zero mean Gaussian White noise inputs. Results are shown to be in agreement with previous linearization studies and Monte Carlo Simulation.

• Journal of the Korea Safety Management & Science
• /
• v.12 no.1
• /
• pp.199-205
• /
• 2010

The research proposes the three-factor random measurement models for estimating the precision about operator, part, tool, and various measurement environments. The combined model with crossed and nested factors is developed to analyze the approximate F test by degrees of freedom given by Satterthwaite and point estimation of precisions from expected mean square. The model developed in this paper can be extended to the three useful models according to the type of nested designs. The study also provides the three-step procedures to evaluate the measurement precisions using three indexes such as SNR(Signal-To-Noise Ratio), R&R TR(Reproducibility&Repeatability-To-Total Precision Ratio), and PTR(Precision-To-Tolerance Ratio), The procedures include the identification of resolution, the improvement of R&R reduction, and the evaluation of precision effect.

• Communications for Statistical Applications and Methods
• /
• v.29 no.5
• /
• pp.561-589
• /
• 2022

When multiple classifications and regression trees are combined, tree-based ensemble models, such as random forest (RF) and Bayesian additive regression trees (BART), are produced. We compare the model structures and performances of various ensemble models for regression settings in this study. RF learns bootstrapped samples and selects a splitting variable from predictors gathered at each node. The BART model is specified as the sum of trees and is calculated using the Bayesian backfitting algorithm. Throughout the extensive simulation studies, the strengths and drawbacks of the two methods in the presence of missing data, high-dimensional data, or highly correlated data are investigated. In the presence of missing data, BART performs well in general, whereas RF provides adequate coverage. The BART outperforms in high dimensional, highly correlated data. However, in all of the scenarios considered, the RF has a shorter computation time. The performance of the two methods is also compared using two real data sets that represent the aforementioned situations, and the same conclusion is reached.

• Journal of Korean Institute of Industrial Engineers
• /
• v.29 no.2
• /
• pp.135-144
• /
• 2003

Traffic performance in a microcellular system is much more affected by cell dwell time and channel holding time in each cell. Cell dwell time of a call is characterized by its mobility pattern, i.e., stochastic changes of moving speed and direction. Cell dwell time provides important information for other analyses on traffic performance such as channel holding time, handover rate, and the average number of handovers per call. In the next generation mobile communication system, the cell size is expected to be much smaller than that of current one to accommodate the increase of user demand and to achieve high bandwidth utilization. As the cell size gets small, traffic performance is much more affected by variable mobility of users, especially by that of pedestrians. In previous work, analytical models are based on simple probability models. They provide sufficient accuracy in a simple second-generation cellular system. However, the role of them is becoming invalid in a picocellular environment where there are rapid change of network traffic conditions and highly random mobility of pedestrians. Unlike in previous work, we propose an improved probability model evolved from so-called Random walk model in order to mathematically formulate variable mobility of pedestrians and analyze the traffic performance. With our model, we can figure out variable characteristics of pedestrian mobility with stochastic correlation. The above-mentioned traffic performance measures are analyzed using our model.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.8
• /
• pp.3656-3671
• /
• 2016

Most of the research in the area of wireless communications exclusively relies on simulations. Further, it is essential that the mobility management strategies and routing protocols should be validated under realistic conditions. Most appropriate mobility models play a pivotal role to determine, whether there is any subtle error or flaw in a proposed model. Simulators are the standard tool to evaluate the performance of mobility models however sometimes they suffer from numerous documented problems. To accomplish the widely acknowledged lack of formalization in this domain, a Coloured Petri nets (CPNs) based random direction mobility model for specification, analysis and validation is presented in this paper for wireless communications. The proposed model does not suffer from any border effect or speed decay issues. It is important to mention that capturing the mobility patterns through CPN is challenging task in this type of the research. Further, an appropriate formalism of CPNs supported to analyze the future system dynamic status. Finally the formal model is evaluated with the state space analysis to show how predefined behavioral properties can be applied. In addition, proposed model is evaluated based on generated simulations to track origins of errors during debugging.

• Proceedings of the Korean Society of Soil and Groundwater Environment Conference
• /
• 2000.11a
• /
• pp.69-73
• /
• 2000

We systematically analyze the influence of fracture junction, solute transfer characteristics on transport patterns in discrete, two-dimensional fracture network models. Regular lattices and random fracture networks with power-law length distributions are considered in conjunction with particle tracking methods. Solute transfer probabilities at fracture junctions are determined from analytical considerations and from simple complete mixing and streamline routing models. For regular fracture networks, mixing conditions at fracture junctions are always dominated by either complete mixing or streamline routing end member cases. Moreover bulk transport properties such as the spreading and the dilution of solute are highly sensitive to the mixing rule. However in power-law length networks there is no significant difference in bulk transport properties, as calculated by assuming either of the two extreme mixing rules. This apparent discrepancy between the effects of mixing properties at fracture junctions in regular and random fracture networks is explained by the statistics of the coordination number and of the flow conditions at fracture intersections. We suggest that the influence of mixing rules on bulk solute transport could be important in systematic orthogonal fracture networks but insignificant in random networks.