• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.48 no.4
• /
• pp.86-97
• /
• 2011

This paper presents a novel algorithm to detect more accurate road lane with image sensor-based color classification and directional edge clustering. With treatment of road region and lane as a recognizable color object, the classification of color cues is processed by an iterative optimization of statistical parameters to each color object. These clustered color objects are taken into considerations as initial kernel information for color object detection and recognition. In order to improve the limitation of object classification using the color cues, the directional edge cures within the estimated region of interest in the lane boundary (ROI-LB) are clustered and combined. The results of color classification and directional edge clustering are optimally integrated to obtain the best detection of road lane. The characteristic of the proposed system is to obtain robust result to all real road environments because of using non-parametric approach based only on information of color and edge clustering without a particular mathematical road and lane model. The experimental results to the various real road environments and imaging conditions are presented to evaluate the effectiveness of the proposed method.

• Applied Chemistry for Engineering
• /
• v.24 no.3
• /
• pp.299-304
• /
• 2013

A 17-run Box-Behnken design was used to optimize the synthesis conditions of a monolithic sorbent. The effects of the amount of monomer (mL), crosslink (mL) and porogen (mL) were investigated. The experimental data were fitted to a second-order polynomial equation by the multiple regression analysis and examined using statistical methods. The adjusted coefficient of determination ($R^2$) of the model was 0.9915. The probability value (p < 0.0001) demonstrated a high significance for the regression model. A mean amount of polymer as 2120.15 mg was produced under the following optimum synthesis conditions: the optimized volumes of monomer, crosslink and porogen are 0.30, 1.40, and 1.47 mL, respectively. This was in good agreement with the predicted model value.

• The Korean Journal of Applied Statistics
• /
• v.21 no.3
• /
• pp.441-453
• /
• 2008

We analyzed Korean professional basketball and baseball players salary under the assumption that it depends on the personal records and contribution to the team in the previous year. We extensively used data visualization tools to check the relationship among the variables, to find outliers and to do model diagnostics. We used multiple linear regression and regression tree to fit the model and used cross-validation to find an optimal model. We check the relationship between variables carefully and chose a set of variables for the stepwise regression instead of using all variables. We found that points per game, number of assists, number of free throw successes, career are important variables for the basketball players. For the baseball pitchers, career, number of strike-outs per 9 innings, ERA, number of homeruns are important variables. For the baseball hitters, career, number of hits, FA are important variables.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.6
• /
• pp.1245-1255
• /
• 2017

The online learning is a process of obtaining the solution for a given objective function where the data is accumulated in real time or in batch units. The stochastic gradient descent method is one of the most widely used for the online learning. This method is not only easy to implement, but also has good properties of the solution under the assumption that the generating model of data is homogeneous. However, the stochastic gradient method could severely mislead the online-learning when the homogeneity is actually violated. We assume that there are two heterogeneous generating models in the observation, and propose the a new stochastic gradient method that mitigate the problem of the heterogeneous models. We introduce a robust mini-batch optimization method using statistical tests and investigate the convergence radius of the solution in the proposed method. Moreover, the theoretical results are confirmed by the numerical simulations.

• Preventive Nutrition and Food Science
• /
• v.13 no.4
• /
• pp.334-339
• /
• 2008

This study was designed to determine the optimum conditions of heating temperature, heating time and dilution rates for producing heated garlic juice by using central composite design of response surface methodology. Garlic 3Department of Biosystems Engineering, Chungbuk National University, Cheongju 361-763, Korea was heated using a high temperature and pressure treatment apparatus. Total soluble solid contents ranged from 4.4 at $130^{\circ}C$ 3 hr and 6 fold dilution to 5.89 at $115^{\circ}C$ for 2 hr and 8 fold dilution. The highest total acidity was 0.55% at $120^{\circ}C$ for 3 hr and 2 fold dilution. The pH ranged from 4.01 at $130^{\circ}C$, 3 hr and 6 fold dilution to 5.85 at $120^{\circ}C$, 1 hr and 6 fold dilution. From the results of statistical analysis on the sensory evaluation the predicted optimum processing conditions for best color, taste, flavor and overall acceptance were $119.41^{\circ}C$, 3.11 hr, 5.85 fold dilution, $118.23^{\circ}C$, 3.57 hr, 3.15 fold dilution, $120.54^{\circ}C$, 3.47 hr, 5.01 fold dilution and $119.54^{\circ}C$, 3.18 hr, 5.66 fold dilution, respectively. The application of response surface methodology for preparing heated garlic juice processing showed a good correlation with high significance.

• Wind and Structures
• /
• v.25 no.2
• /
• pp.101-129
• /
• 2017

Local transient extreme wind loads caused by group tower-related interference are among the major reasons that lead to wind-induced damage of super-large cooling towers. Four-tower arrangements are the most commonly seen patterns for super-large cooling towers. We considered five typical four-tower arrangements in engineering practice, namely, single row, rectangular, rhombic, L-shaped, and oblique L-shaped. Wind tunnel tests for rigid body were performed to determine the influence of different arrangements on static and dynamic wind loads and extreme interference effect. The most unfavorable working conditions (i.e., the largest overall wind loads) were determined based on the overall aerodynamic coefficient under different four-tower arrangements. Then we calculated the one-, two- and three-dimensional aerodynamic loads under different four-tower arrangements. Statistical analyses were performed on the wind pressure signals in the amplitude and time domains under the most unfavorable working conditions. On this basis, the non-Gaussian distribution characteristics of aerodynamic loads on the surface of the cooling towers under different four-tower arrangements were analyzed. We applied the Sadek-Simiu procedure to the calculation of two- and three-dimensional aerodynamic loads in the cooling towers under the four-tower arrangements, and the extreme wind load distribution patterns under the most unfavorable working conditions in each arrangement were compared. Finally, we proposed a uniform equation for fitting the extreme wind loads under the four-tower arrangements; the accuracy and reliability of the equation were verified. Our research findings will contribute to the optimization of the four-tower arrangements and the determination of extreme wind loads of super-large cooling towers.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.44 no.1
• /
• pp.31-36
• /
• 2011

This study developed a normal temperature-circulatable retort-sterilized crab analog (RSCA) with a long shelf-life by retort sterilization. We examined the optimum sterilization conditions and quality characteristics of the RSCA. A central composite design for response surface methodology (RSM) was adopted to optimize of the retort sterilization. The central composite design consisted of 11 samples: four factorial points, four star points, and three central points. The sterilization temperature and $F_0$-value for retort sterilization were the independent variables, and the shearing force, whiteness, and sensory score for texture were the dependent variable. The RSM analysis of the multiple response optimization for the RSCA using the Minitap statistical programing gave an $F_0$-value of 3.3 min. at $117.5^{\circ}C$. The RSM analysis also indicated that the sterilization temperature during retort sterilization was the most influential factor, while the $F_0$-value had little effect on the quality of the RSCA. The moisture, pH and volatile basic nitrogen contents of the RSCA product were 73.1%, 7.3 and 17.2 mg/100 g, respectively, and viable cells count was >18 CFU/g. And sensory quality characteristics of this RSCA was similar to those of a commercial crab analog.

• Journal of KIISE:Computing Practices and Letters
• /
• v.14 no.5
• /
• pp.527-531
• /
• 2008

For statistical modeling, the model parameters are usually estimated by maximizing a probability measure, such as the likelihood or the posterior. In contrast, a variational Bayesian method treats the parameters of a model as probability distributions and computes optimal distributions for them rather than values. It has been shown that this approach effectively avoids the overfitting problem, which is common with other parameter optimization methods. This paper applies a variational Bayesian technique to surface fitting for height field data. Then, we propose point cloud denoising based on the basic surface fitting technique. Validation experiments and further tests with scan data verify the robustness of the proposed method.

• Mycobiology
• /
• v.47 no.4
• /
• pp.512-520
• /
• 2019

Statistical experimental methods were used to optimize the medium for mass production of a novel laccase3 (Lac3) by recombinant Saccharomyces cerevisiae TYEGLAC3-1. The basic medium was composed of glucose, casamino acids, yeast nitrogen base without amino acids (YNB w/o AA), tryptophan, and adenine. A one-factor-at-a-time approach followed by the fractional factorial design identified galactose, glutamic acid, and ammonium sulfate, as significant carbon, nitrogen, and mineral sources, respectively. The steepest ascent method and response surface methodology (RSM) determined that the optimal medium was (g/L): galactose, 19.16; glutamic acid, 5.0; and YNB w/o AA, 10.46. In this medium, the Lac3 activity (277.04 mU/mL) was 13.5 times higher than that of the basic medium (20.50 mU/mL). The effect of temperature, pH, agitation (rpm), and aeration (vvm) was further examined in a batch fermenter. The best Lac3 activity was 1176.04 mU/mL at 25 ℃, pH 3.5, 100 rpm, and 1 vvm in batch culture.

• Korean Chemical Engineering Research
• /
• v.51 no.6
• /
• pp.677-684
• /
• 2013

A mixed refrigerant cycle (MRC) has been widely used in liquefaction of natural gas because it is simple and easily operable with reasonable equipment costs. One of the important techniques in MRC is selection of a refrigerant mixture and decision of its optimum mixing ratio. In this work, it is examined whether mixture components (refrigerants) and their mixing ratio influence performance of general MRC processes. In doing this, mixture design and response surface method, which are well-known statistical techniques, are used to find optimal mixture refrigerants and their optimal mixing ratio that minimize total energy consumption of the entire liquefaction process. A MRC process using several refrigerants and various mixing ratios is simulated by Aspen HYSYS and mixture design and response surface method are implemented using Minitab. According to the results, methane ($C_1$), ethane ($C_2$), propane ($C_3$) and nitrogen ($N_2$) are selected as best mixture refrigerants and the determined mixture ratio (mole ration) can reduce total energy consumption by up to 50%.