• International Journal of Control, Automation, and Systems
• /
• v.6 no.1
• /
• pp.109-118
• /
• 2008

We present two estimators for discrete non-Gaussian and nonstationary probability density estimation based on a dynamic Bayesian network (DBN). The first estimator is for off line computation and consists of a DBN whose transition distribution is represented in terms of kernel functions. The estimator parameters are the weights and shifts of the kernel functions. The parameters are determined through a recursive learning algorithm using maximum likelihood (ML) estimation. The second estimator is a DBN whose parameters form the transition probabilities. We use an asymptotically convergent, recursive, on-line algorithm to update the parameters using observation data. The DBN calculates the state probabilities using the estimated parameters. We provide examples that demonstrate the usefulness and simplicity of the two proposed estimators.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.48 no.6
• /
• pp.124-132
• /
• 2011

In a modern electronic warfare, radar signal environments become more denser and complex. Therefor the capability of reliable signal analysis techniques is required for ES(Electronic warfare Support) system to identify and analysis individual emitter signals from received signals. In this paper, we propose the new signal grouping algorithm to ensure the reliable signal analysis and to reduce the cost of the signal processing steps in the ES. The proposed grouping algorithm uses KDE(Kernel Density Estimator) and its CDF(Cumulative Distribution Function) to compose windows considering the statistical distribution characteristics based on the radar frequency modulation type. Simulation results show the good performance of the proposed technique in the signal grouping.

• Journal of Biosystems Engineering
• /
• v.6 no.1
• /
• pp.73-82
• /
• 1981

The physical properties of grain are very important for the design of handling, sorting, processing, and storage system. On the physical properties of grain, volume, bulk density, true density, specific gravity, and porosity arc the major factors affecting the thermal properties of grain. This study was conducted to determine experimentally the above physical properties of rough rice (3 Japonica-type, 3 Indica-type) and barley (covered, naked) as a function of moisture content ranged from about 10% to 25% (w.b). The results of this study are summarized as follows; 1. The volume of grain kernel increased with moisture content for both rice and barley. The volume of those grain kernel was in the range of $2.2068{\times}10^{-8}{\sim}3.3960{\times}10^{-8}m^3$ at the moisture content of 14%. 2. The bulk density of rice increased linearly with moisture content for Japonica-type rough rice and quadratically for Indica-type rough rice, but the bulk density of barley decreased linearly with moisture content. The bulk density of the grain was in the range of 501.14~689.13kg/$m^3$ at the moisture content of 14%. 3. The true density of whole grain decreased linearly with moisture content, and was in the range of 1019.49~1139.75kg/$m^3$ at the moisture content of 14%. 4. The porosity of rice decreased linearly with moisture content for Japonica-type rough rice and quadratically for Indica-type rough rice, but the porosity of barley increased linearly with moisture content. The porosity of the grain was in the range of 39.51~50.83% at the moisture content of 14%. 5. The regression equations of the physical properties such as volume, bulk density, true density, and porosity of the grain were determined as a function of moisture content.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.28 no.2B
• /
• pp.225-236
• /
• 2008

The techniques to calculate the probability precipitation for the design of hydrological projects can be determined by the point frequency analysis and the regional frequency analysis. Probability precipitation usually calculated by point frequency analysis using rainfall data that is observed in rainfall observatory which is situated in the basin. Therefore, Probability precipitation through point frequency analysis need observed rainfall data for enough periods. But, lacking precipitation data can be calculated to wrong parameters. Consequently, the regional frequency analysis can supplement the lacking precipitation data. Therefore, the regional frequency analysis has weaknesses compared to point frequency analysis because of suppositions about probability distributions. In this paper, rainfall observatory in Korea did grouping by cluster analysis using position of timely precipitation observatory and characteristic time rainfall. Discordancy and heterogeneity measures verified the grouping precipitation observatory by the cluster analysis. So, there divided rainfall observatory in Korea to 6 areas, and the regional frequency analysis applies index-flood techniques and L-moment techniques. Also, the probability precipitation was calculated by the regional frequency analysis using variable kernel density function. At the results, the regional frequency analysis of the variable kernel function can utilize for decision difficulty of suitable probability distribution in other methods.

• Journal of Korea Multimedia Society
• /
• v.19 no.3
• /
• pp.633-641
• /
• 2016

Automatic defect inspection system is composed of the step in the pre-processing, defect candidate detection, and classification. Polarizing films containing various defects should be minimized over-detection for classifying defect blobs. In this paper, we propose a defect detection algorithm using a skewness of histogram for minimizing over-detection. In order to detect up defects with similar to background pixel, we are used the characteristics of the local region. And the real defect pixels are distinguished from the noise using the probability density function. Experimental results demonstrated the minimized over-detection by utilizing the artificial images and real polarizing film images.

• Proceedings of the IEEK Conference
• /
• 2008.06a
• /
• pp.973-974
• /
• 2008

As radar signal environments become denser and more complex, the capability of high-speed and accurate signal analysis is required for ES(Electronic warfare Support) system to identify individual radar signals at real-time. In this paper, we propose the new novel clustering algorithm of radar pulses to alleviate the load of signal analysis process and support reliable analysis. The proposed algorithm uses KDE(Kernel Density Estimation) and its CDF(Cumulative Distribution Function) to compose clusters considering the distribution characteristics of pulses. Simulation results show the good performance of the proposed clustering algorithm in clustering and classifying the emitters.

• Journal of the Korean Statistical Society
• /
• v.34 no.1
• /
• pp.49-60
• /
• 2005

In this paper we introduce a nonparametric method for estimating the probability density function of detection distances in line transect sampling. The estimator is obtained using a frequency histogram density estimation method. The asymptotic properties of the proposed estimator are derived and compared with those of the kernel estimator under the assumption that the data collected satisfy the shoulder condition. We found that the asymptotic mean square error (AMSE) of the two estimators have about the same convergence rate. The formula for the optimal histogram bin width is derived which minimizes AMSE. Moreover, the performances of the corresponding k-nearest-neighbor estimators are studied through simulation techniques. In the absence of our knowledge whether the shoulder condition is valid or not a new semi-parametric model is suggested to fit the line transect data. The performances of the proposed two estimators are studied and compared with some existing nonparametric and semiparametric estimators using simulation techniques. The results demonstrate the superiority of the new estimators in most cases considered.

• Journal of the Korean Data and Information Science Society
• /
• v.22 no.1
• /
• pp.107-113
• /
• 2011

We present methods for studying the log-density ratio, which allow us to select which predictors are needed, and how they should be included in the logistic regression model. Under multivariate normal distributional assumptions, we investigate the form of the log-density ratio as a function of many predictors. The linear, quadratic and crossproduct terms are required in general. If two covariance matrices are equal, then the crossproduct and quadratic terms are not needed. If the variables are uncorrelated, we do not need the crossproduct terms, but we still need the linear and quadratic terms.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.32 no.3
• /
• pp.173-181
• /
• 2019

To estimate probabilistic distribution function from experimental data, kernel density estimation(KDE) is mostly used in cases when data is insufficient. The estimated distribution using KDE depends on bandwidth selectors that smoothen or overfit a kernel estimator to experimental data. In this study, various bandwidth selectors such as the Silverman's rule of thumb, rule using adaptive estimates, and oversmoothing rule, were compared for accuracy and conservativeness. For this, statistical simulations were carried out using assumed true models including unimodal and multimodal distributions, and, accuracies and conservativeness of estimating distribution functions were compared according to various data. In addition, it was verified how the estimated distributions using KDE with different bandwidth selectors affect reliability analysis results through simple reliability examples.

• Environmental and Resource Economics Review
• /
• v.18 no.1
• /
• pp.105-138
• /
• 2009

This study investigates the spatial distribution of the degree of water pollutants in Han River using data obtained by the water pollution observation stations. This study estimates a non -parametric kernel density function for each water pollutants, and tests a significant difference between two estimated distribution functions. Next, Generalized Entropy inequality indices are evaluated and this research tests difference of inequality indices between two years using bootstrapping method. Lastly in a dynamic of view, it is analyzed that there are significant changes in the ranking of water pollution level. Estimation results show that the degree of inequality in spatial distribution of water pollution tends to be stable or decreasing for last 15 years in a great part of water pollutants, and ranking of water pollution level hardly changes in Han River.