• Journal of Korea Water Resources Association
• /
• v.55 no.3
• /
• pp.229-243
• /
• 2022

The effects of monthly meteorological data measured at 11 stations in South Korea on pan coefficient were analyzed to develop the four types of multiple linear regression models for estimating pan coefficients. To evaluate the applicability of developed models, the models were compared with six previous models. Pan coefficients were most affected by air temperature for January, February, March, July, November and December, and by solar radiation for other months. On the whole, for 12 months of the year, the effects of wind speed and relative humidity on pan coefficient were less significant, compared with those of air temperature and solar radiation. For all meteorological stations and months, the model developed by applying 5 independent variables (wind speed, relative humidity, air temperature, ratio of sunshine duration and daylight duration, and solar radiation) for each station was the most effective for evaporation estimation. The model validation results indicate that the multiple linear regression models can be applied to some particular stations and months.

• Sleep Medicine and Psychophysiology
• /
• v.15 no.1
• /
• pp.5-11
• /
• 2008

Sleep is an essential process maintaining the life cycle of the human. In parallel with physiological, cognitive, subjective, and behavioral changes that take place during the sleep, there are remarkable changes in the electroencephalogram (EEG) that reflect the underlying electro-physiological activity of the brain. However, analyzing EEG and relating the results to clinical observations is often very hard due to the complexity and a huge data amount. In this article, I introduce several linear and non-linear tools, developed to analyze a huge time series data in many scientific researches, and apply them to EEG to characterize various sleep states. In particular, the spectral analysis, detrended fluctuation analysis (DFA), and synchrony analysis are administered to EEG recorded during nocturnal polysomnography (NPSG) processes and daytime multiple sleep latency tests (MSLT). I report that 1) sleep stages could be differentiated by the spectral analysis and the DFA ; 2) the gradual transition from Wake to Sleep during the sleep onset could be illustrated by the spectral analysis and the DFA ; 3) electrophysiological properties of narcolepsy could be characterized by the DFA ; 4) hypnic jerks (sleep starts) could be quantified by the synchrony analysis.

• Journal of Korea Multimedia Society
• /
• v.20 no.8
• /
• pp.1406-1420
• /
• 2017

Artificial neural networks (ANNs) play an important role in the fields of function approximation, prediction, and classification. ANN performance is critically dependent on the input parameters, including the number of neurons in each layer, and the optimal values of weights and biases assigned to each neuron. In this study, we apply the particle swarm optimization method, a popular optimization algorithm for determining the optimal values of weights and biases for every neuron in different layers of the ANN. Several regression models, including general linear regression, Fourier regression, smoothing spline, and polynomial regression, are conducted to evaluate the proposed method's prediction power compared to multiple linear regression (MLR) methods. In addition, residual analysis is conducted to evaluate the optimized ANN accuracy for both training and test datasets. The experimental results demonstrate that the proposed method can effectively determine optimal values for neuron weights and biases, and high accuracy results are obtained for prediction applications. Evaluations of the proposed method reveal that it can be used for prediction and estimation purposes, with a high accuracy ratio, and the designed model provides a reliable technique for optimization. The simulation results show that the optimized ANN exhibits superior performance to MLR for prediction purposes.

• Structural Engineering and Mechanics
• /
• v.61 no.2
• /
• pp.193-207
• /
• 2017

In this study, the vibration of an electrostatically actuated micro cantilever beam is analyzed in which a viscoelastic layer covers a portion of the micro beam length. This proposed model is considered as the main element of mass and pollutant micro sensors. The nonlinear motion equation is extracted by means of Hamilton principle, considering nonlinear shortening effect for Euler-Bernoulli beam. The non-linear effects of electrostatic excitation, geometry and inertia have been taken into account. The viscoelastic model is assumed as Kelvin-Voigt model. The motion equation is discretized by Galerkin approach. The linear free vibration mode shapes of non-uniform micro beam i.e. the linear mode shape of the system by considering the geometric and inertia effects of viscoelastic layer, have been employed as comparison function in the process of the motion equation discretization. The discretized equation of motion is solved by the use of multiple scale method of perturbation theory and the results are compared with the results of numerical Runge-Kutta approach. The frequency response variations for different lengths and thicknesses of the viscoelastic layer have been founded. The results indicate that if a constant volume of viscoelastic layer is to be deposited on the micro beam for mass or gas sensor applications, then a modified configuration may be found by using the analysis of this paper.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2003.10a
• /
• pp.59-65
• /
• 2003

Four kinds of $Si_3N_4$-based ceramic cutting tools with different sintering time were fabricated to investigante the effect of sintering time on the microstructure, mechanical properties, grain sizes and the cutting performance. An endeavor was also made to determine the relation among mechanical property, Brain size and tool life. $Si_3N_4$ home made cutting tool sintered for 1 hour under $1760^{\circ}$ temperature and 25MPa pressure showed the best cutting performance among selected ceramic tools during machining both Bray cast iron and heat treated SCM440. Multiple linear regression model was used to estimate the tool lift from mechanical property, grain size and showed good result. It was also shown that hardness imposed the biggest offect on tool life.

• Environmental Analysis Health and Toxicology
• /
• v.15 no.1_2
• /
• pp.39-43
• /
• 2000

Linear alkyl sulfonates (LAS) constitute a large fraction of the surfactants used in cleaning processes in households, trade and industry Despite the industrial significance and the possible environmental impact of these compounds, the fast and inexpensive determination of LAS concentrations is still a difficult task. In this study, near infrared (NIR) spectroscopy which is a rapid spectroscopic analysis method compared with a traditional analytical method for the measurement of LAS concentration such as HPLC, GC and standard wet chemistry method. NIR spectra of LAS between 0.313 and 25.0% (w/v) in water were utilized to develop a calibration model. The best results (R ＝ 0.998, SEP ＝ 0.244% (w/v)) obtained by using partial least-squares regression with spectral data treatment and 2nd derivatization were comparable to the results (SEC ＝ 0.186% (w/v), SEP ＝ 0.206% (w/v)) obtained by using multiple linear least-squares regression (MLR). However, models based on derivative spectra have no significant advantage with MLR.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.20 no.4
• /
• pp.65-76
• /
• 2017

The purpose of this study was to present a prediction model that reflects crime risk area analysis, including factors and spatial characteristics, as a precursor to preparing an alternative plan for crime prevention and design. This analysis of criminal cases in high-risk areas revealed clusters in which approximately 25% of the cases within the study area occurred, distributed evenly throughout the region. This means that using a multiple linear regression model might overestimate the crime rate in some regions and underestimate in others. It also suggests that the number of deserted houses in an analyzed region has a negative relationship with the dependent variable, based on the multiple linear regression model results, and can also have different influences depending on the region. These results reveal that closure signs in a study area affect the dependent variable differently, depending on the region, rather than a simple or direct relationship with the dependent variable, as indicated by the results of the multiple linear regression model.

• Journal of Korean Society for Quality Management
• /
• v.48 no.4
• /
• pp.535-552
• /
• 2020

Purpose: The main theme of this study is to derive a specific quality loss function with multiple characteristics according to the same analytical structure as the single characteristic quality loss function of Taguchi. In other words, it presents an analytical framework for measuring quality costs that can be controlled in practice. Methods: This study followed the analytical methodology through geometric, linear algebraic, and statistical approaches Results: The function suggested by this study is as follows; $$L(x_1,x_2,{\cdots},x_t)={\sum\limits_{i=1}^{t}}k_i\{x_i+{\sum\limits_{j=1}^{t}}\({\rho}_{ij}{\frac{d_i}{d_j}}\)x_j\}x_i$$ Conclusion: This paper derived the quality loss function with multiple quality characteristics to expand the usefulness of the Taguchi quality loss function. The function derived in this paper would be more meaningful to estimate quality costs under the practical situation and general structure with multiple quality characteristics than the function by linear algebraic approach in response surface analysis.

• International Journal of Computer Science & Network Security
• /
• v.23 no.11
• /
• pp.21-31
• /
• 2023

In this paper, we focused on the problem of evaluating multi-class classification accuracy and simulation of multiple classifier performance metrics. Multi-class classifiers for sentiment analysis involved many challenges, whereas previous research narrowed to the binary classification model since it provides higher accuracy when dealing with text data. Thus, we take inspiration from the non-linear Support Vector Machine to modify the algorithm by embedding dynamic hyperplanes representing multiple class labels. Then we analyzed the performance of multi-class classifiers using macro-accuracy, micro-accuracy and several other metrics to justify the significance of our algorithm enhancement. Furthermore, we hybridized Enhanced Convolution Neural Network (ECNN) with Dynamic Support Vector Machine (DSVM) to demonstrate the effectiveness and efficiency of the classifier towards multi-class text data. We performed experiments on three hybrid classifiers, which are ECNN with Binary SVM (ECNN-BSVM), and ECNN with linear Multi-Class SVM (ECNN-MCSVM) and our proposed algorithm (ECNNDSVM). Comparative experiments of hybrid algorithms yielded 85.12 % for single metric accuracy; 86.95 % for multiple metrics on average. As for our modified algorithm of the ECNN-DSVM classifier, we reached 98.29 % micro-accuracy results with an f-score value of 98 % at most. For the future direction of this research, we are aiming for hyperplane optimization analysis.

• Journal of Mechanical Science and Technology
• /
• v.16 no.6
• /
• pp.819-828
• /
• 2002

An analytical method is presented for evaluation of the steady state periodic behavior of nonlinear systems. This method is based on the substructure synthesis formulation and a MS (multiple scales) procedure, which is applied to the analysis of nonlinear responses. The proposed procedure reduces the size of large degrees-of-freedom problem in solving nonlinear equations. Feasibility and advantages of the proposed method are illustrated with the nonlinear rotating machine system as an example of large mechanical structure systems. In addition, its efficiency for nonlinear response prediction will be shown by comparison of other conventional methods.