• Communications for Statistical Applications and Methods
• /
• v.23 no.4
• /
• pp.355-362
• /
• 2016

Ensemble methods often help increase prediction ability in various predictive models by combining multiple weak learners and reducing the variability of the final predictive model. In this work, we demonstrate that ensemble methods also enhance the accuracy of prediction under kernel ridge regression and kernel logistic regression classification. Here we apply bagging and random forests to two kernel-based predictive models; and present the procedure of how bagging and random forests can be embedded in kernel-based predictive models. Our proposals are tested under numerous synthetic and real datasets; subsequently, they are compared with plain kernel-based predictive models and their subsampling approach. Numerical studies demonstrate that ensemble approach outperforms plain kernel-based predictive models.

• The Journal of the Korea Contents Association
• /
• v.10 no.6
• /
• pp.134-143
• /
• 2010

This paper proposes a framework for classifying metric-based software quality prediction models, especially case of software criticality, into four types. Models are classified along two vectors: input metric forms and the necessity of past project data. Each type has its own characteristics and its strength and weakness are compared with those of other types using newly defined criteria. Through this qualitative evaluation each organization can choose a proper model to suit its environment. My earlier studies of criticality prediction model implemented specific models in each type and evaluated their prediction performances. In this paper I analyze the experimental results and show that the characteristics of a model type is the another key of successful model selection.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2008.03a
• /
• pp.587-598
• /
• 2008

In the trackbed design using elastic multi-layer model, the stress-dependent resilient modulus is an important input parameter, which reflects substructure performance under repeated traffic loading. The resilient moduli of crushed stone and weathered granite soil were developed using nonlinear dynamic stiffness, which can be measured by in-situ and laboratory seismic tests. The prediction models of resilient modulus varying with the deviatoric or bulk stress were proposed (Park et al., 2008). To investigate the performance of the prediction models proposed herein, the elastic response of the test trackbed near PyeongTaek, Korea was evaluated using a 3-D nonlinear elastic computer program (GEOTRACK) and compared with measured elastic vertical displacement during the passages of freight and passenger trains. The material types of the test sub-ballasts are crushed stone and weathered granite soil, respectively. The calculated vertical displacements within the sub-ballasts are within the order of 1mm, and agree well with measured values with the reasonable margin. The prediction models are thus concluded to work properly in the preliminary investigation. The prediction models proposed for resilient modulus were verified by the comparison of the calculated vertical displacements with measured ones during train passages.

• Journal of ILASS-Korea
• /
• v.12 no.3
• /
• pp.146-153
• /
• 2007

The prediction of diesel spray penetration has been the subject of several works and intensive investigations are still underway by many researchers. It is required to summarize the correlations developed before 1990 days and to introduce the correlations reported recently in the literature. The existing zero-dimensional models for the prediction of diesel fuel spray penetration can be classified as theoretical and empirical correlations. Of various correlations, the models considered in this paper were selected as based on the evaluation results of previous reviews and the recently published works in the literature. The existing theoretical correlations can be classified into seven categories and the existing empirical ones as two categories in this review. According to the review of existing models, the dominating factors for the prediction of spray tip penetration are the spray angle, discharge coefficient, pressure drop across nozzle, ambient density and orifice diameter and time after the start of injection. Especially, the definition for the measurement of spray angle is different with researchers. It is required to evaluate the existing spray tip penetration models for the very high injection pressure and other fuel sprays such as DME. It is also required to evaluate the correlations for the prediction of diesel spray penetration with the connection of liquid-phase penetration.

• Computers and Concrete
• /
• v.22 no.2
• /
• pp.197-207
• /
• 2018

In the present study, group method of data handling networks (GMDH) are adopted and evaluated for shear strength prediction of both FRP-reinforced concrete members with and without stirrups. Input parameters considered for the GMDH are altogether 12 influential geometrical and mechanical parameters. Two available and very recently collected comprehensive datasets containing 112 and 175 data samples are used to develop new models for two cases with and without shear reinforcement, respectively. The proposed GMDH models are compared with several codes of practice. An artificial neural network (ANN) model and an ANFIS based model are also developed using the same databases to further assessment of GMDH. The accuracy of the developed models is evaluated by statistical error parameters. The results show that the GMDH outperforms other models and successfully can be used as a practical and effective tool for shear strength prediction of members without stirrups ($R^2=0.94$) and with stirrups ($R^2=0.95$). Furthermore, the relative importance and influence of input parameters in the prediction of shear capacity of reinforced concrete members are evaluated through parametric and sensitivity analyses.

• Steel and Composite Structures
• /
• v.28 no.2
• /
• pp.251-266
• /
• 2018

Compared to conventional flat web I-beams, the prediction of shear buckling stress of corrugated web steel beams (CWSBs) is not straightforward. But the CWSBs combined advantages of lightweight large spans with low-depth high load-bearing capacities justify dealing with such difficulties. This work investigates experimentally and analytically the shear strength of trapezoidal CWSBs. A set of large scale CWSBs are manufactured and tested to failure in shear. The results are compared with widely accepted CWSBs shear strength prediction models. Confirmed by the experimental results, the linear buckling analyses of trapezoidal corrugated webs demonstrated that the local shear buckling occurs only in the flat plane folds of the web, while the global shear buckling occurs over multiple folds of the web. New analytical prediction model accounting for the interaction between the local and global shear buckling of CWSBs is proposed. Experimental results from the current work and previous studies are compared with the proposed analytical prediction model. The predictions of the proposed model are significantly better than all other studied models. In light of the dispersion of test data, accuracy, consistency, and economical aspects of the prediction models, the authors recommend their proposed model for the design of CWSBs over the rest of the models.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.28 no.11
• /
• pp.1813-1821
• /
• 2004

Until now, the 40% of wall thickness criterion, which is generally used for the plugging of steam generator tubes, has been applied only to a single cracked geometry. In the previous study by the authors, a total number of 9 local failure prediction models were introduced to estimate the coalescence load of two collinear through-wall cracks and, then, the reaction force model and plastic zone contact model were selected as the optimum ones. The objective of this study is to estimate the coalescence load of two collinear through-wall cracks in steam generator tube by using the optimum local failure prediction models. In order to investigate the applicability of the optimum local failure prediction models, a series of plastic collapse tests and corresponding finite element analyses for two collinear through-wall cracks in steam generator tube were carried out. Thereby, the applicability of the optimum local failure prediction models was verified and, finally, a coalescence evaluation diagram which can be used to determine whether the adjacent cracks detected by NDE coalesce or not has been developed.

• Journal of the Korean Solar Energy Society
• /
• v.38 no.1
• /
• pp.1-11
• /
• 2018

In this work, we propose weather prediction models to estimate hourly outdoor temperatures and solar irradiance in the next day using forecasting information. Hourly weather data predicted by the proposed models are useful for setting system operating strategies for the next day. The outside temperature prediction model considers 3-hourly temperatures forecasted by Korea Meteorological Administration. Hourly data are obtained by a simple interpolation scheme. The solar irradiance prediction is achieved by constructing a dataset with the observed cloudiness and correspondent solar irradiance during the last two weeks and then by matching the forecasted cloud factor for the next day with the solar irradiance values in the dataset. To verify the usefulness of the weather prediction models in predicting a short-term building load, the predicted data are inputted to a TRNSYS building model, and results are compared with a reference case. Results show that the test case can meet the acceptance error level defined by the ASHRAE guideline showing 8.8% in CVRMSE in spite of some inaccurate predictions for hourly weather data.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.55 no.4
• /
• pp.1-11
• /
• 2013

Unsaturated hydraulic conductivity function is one of key parameters to solve the flow phenomena in problems of landslide. Prediction models for hydraulic conductivity function related to soil-water retention curve equations in many geotechnical applications have been still used instead of direct measurement of the hydraulic conductivity function since prediction models from soil-water retention curve equations are attractive for their fast and easy use and low cost. However, many researchers found that prediction models for the hydraulic conductivity function can not predict the hydraulic conductivity exactly in comparison with experimental outputs. This research introduced an inverse analysis to evaluate the hydraulic conductivity function corresponding to experimental output from the flow pump system. Optimisation process was carried out to obtain the hydraulic conductivity function. This research showed that the inverse analysis with flow pump system was suitable to assess the hydraulic conductivity in unsaturated soil, and the prediction models for the hydraulic conductivity were led to the significant discrepancy from actual experimental outputs.

• Geomechanics and Engineering
• /
• v.33 no.2
• /
• pp.183-194
• /
• 2023

This study presents the prediction of the California bearing ratio (CBR) of coarse- and fine-grained soils using artificial intelligence technology. The group method of data handling (GMDH) algorithm, an artificial neural network-based model, was used in the prediction of the CBR values. In the design of the prediction models, various combinations of independent input variables for both coarse- and fine-grained soils have been used. The results obtained from the designed GMDH-type neural networks (GMDH-type NN) were compared with other regression models, such as linear, support vector, and multilayer perception regression methods. The performance of models was evaluated with a regression coefficient (R²), root-mean-square error (RMSE), and mean absolute error (MAE). The results showed that GMDH-type NN algorithm had higher performance than other regression methods in the prediction of CBR value for coarse- and fine-grained soils. The GMDH model had an R² of 0.938, RMSE of 1.87, and MAE of 1.48 for the input variables {G, S, and MDD} in coarse-grained soils. For fine-grained soils, it had an R² of 0.829, RMSE of 3.02, and MAE of 2.40, when using the input variables {LL, PI, MDD, and OMC}. The performance evaluations revealed that the GMDH-type NN models were effective in predicting CBR values of both coarse- and fine-grained soils.