• International Journal of CAD/CAM
• /
• 제7권1호
• /
• pp.1-10
• /
• 2007

The compression of CAD models is a key technology for realizing Internet-based collaborative product development because big model sizes often prohibit us to achieve a rapid product information transmission. Although there exist some algorithms for compressing discrete CAD models, original precise CAD models are focused on in this paper. Here, the characteristics of hierarchical structures in CAD models and the distribution of their redundant data are exploited for developing a novel data encoding method. In the method, different encoding rules are applied to different types of data. Geometric data is a major concern for reducing model sizes. For geometric data, the control points of B-spline curves and surfaces are compressed with the second-order predictions in a local coordinate system. Based on analysis to the distortion induced by quantization, an efficient method for computation of the distortion is provided. The results indicate that the data size of CAD models can be decreased efficiently after compressed with the proposed method.

• 천문학회보
• /
• 제40권1호
• /
• pp.64.3-64.3
• /
• 2015

We have developed a set of daily solar flare peak flux forecast models for strong flares using multiple linear regression and artificial neural network methods. We consider input parameters as solar activity data from January 1996 to December 2013 such as sunspot area, X-ray flare peak flux and weighted total flux of previous day, and mean flare rates of McIntosh sunspot group (Zpc) and Mount Wilson magnetic classification. For a training data set, we use the same number of 61 events for each C-, M-, and X-class from Jan. 1996 to Dec. 2004, while other previous models use all flares. For a testing data set, we use all flares from Jan. 2005 to Nov. 2013. The best three parameters related to the observed flare peak flux are weighted total flare flux of previous day (r = 0.51), X-ray flare peak flux (r = 0.48), and Mount Wilson magnetic classification (r = 0.47). A comparison between our neural network models and the previous models based on Heidke Skill Score (HSS) shows that our model for X-class flare is much better than the models and that for M-class flares is similar to them. Since all input parameters for our models are easily available, the models can be operated steadily and automatically in near-real time for space weather service.

• 한국측량학회:학술대회논문집
• /
• 한국측량학회 2004년도 춘계학술발표회논문집
• /
• pp.491-500
• /
• 2004

As the applications of 3D GIS rapidly increase, the need for acquisition and continuos update of urban models is increasingly emphasized. Particularly, building models has been considered as the most crucial component of urban models. Many researchers thus have focused on building extraction from mainly aerial photos or airborne LIDAR data but so far mostly failed to achieve satisfactory results in terms of both completeness and precision because of the intrinsic limitation of the sensory data themselves. Therefore, instead of the airborne sensors, we utilize a terrestrial LIDAR to generate precise and complete building models. This paper presents the overview of the sensors for data acquisition, describes data processing methods for building modelling from the acquired data and summerizes the experimental results.

• 대한산업공학회지
• /
• 제42권5호
• /
• pp.327-336
• /
• 2016

Prediction of customer failure rates plays an important role for establishing appropriate management policies and improving the profitability for industries. For these reasons, many LCD (Liquid crystal display) manufacturing industries have attempted to construct prediction models for customer failure rates. However, most traditional models are based on the parametric approaches requiring the assumption that the data follow a certain probability distribution. To address the limitation posed by the distributional assumption underpinning traditional models, we propose using parameter-free data mining models for predicting customer failure rates. In addition, we use various information associated with product attributes and field return for more comprehensive analysis. The effectiveness and applicability of the proposed method were demonstrated with a real dataset from one of the leading LCD companies in South Korea.

• Journal of the Korean Data and Information Science Society
• /
• 제14권3호
• /
• pp.641-651
• /
• 2003

This study suggests the hybrid models for interest rate forecasting using structural changes (or change points). The basic concept of this proposed model is to obtain significant intervals caused by change points, to identify them as the change-point groups, and to reflect them in interest rate forecasting. The model is composed of three phases. The first phase is to detect successive structural changes in the U. S. Treasury bill rate dataset. The second phase is to forecast the change-point groups with data mining classifiers. The final phase is to forecast interest rates with backpropagation neural networks (BPN). Based on this structure, we propose three hybrid models in terms of data mining classifier: (1) multivariate discriminant analysis (MDA)-supported model, (2) case-based reasoning (CBR)-supported model, and (3) BPN-supported model. Subsequently, we compare these models with a neural network model alone and, in addition, determine which of three classifiers (MDA, CBR and BPN) can perform better. For interest rate forecasting, this study then examines the prediction ability of hybrid models to reflect the structural change.

• Asia pacific journal of information systems
• /
• 제26권1호
• /
• pp.66-79
• /
• 2016

The availability of detailed data on customers' online behaviors and advances in big data analysis techniques enable us to predict consumer behaviors. In the past, researchers have built purchase prediction models by analyzing clickstream data; however, these clickstream-based prediction models have had several limitations. In this study, we propose a new method for purchase prediction that combines information theory with machine learning techniques. Clickstreams from 5,000 panel members and data on their purchases of electronics, fashion, and cosmetics products were analyzed. Clickstreams were summarized using the 'entropy' concept from information theory, while 'random forests' method was applied to build prediction models. The results show that prediction accuracy of this new method ranges from 0.56 to 0.83, which is a significant improvement over values for clickstream-based prediction models presented in the past. The results indicate further that consumers' information search behaviors differ significantly across product categories.

• Smart Structures and Systems
• /
• 제15권3호
• /
• pp.751-767
• /
• 2015

A new procedure is proposed for assessing probabilistic condition of structures considering effect of measured data uncertainty. In this procedure, multiple Finite Element (FE) models are identified by using weighting vectors that represent the uncertainty conditions of measured data. The distribution of structural parameters is analysed using a Principal Component Analysis (PCA) in relation to uncertainty conditions, and the identified models are classified into groups according to their similarity by using a K-means method. The condition of a structure is then assessed probabilistically using FE models in the classified groups, each of which represents specific uncertainty condition of measured data. Yeondae bridge, a steel-box girder expressway bridge in Korea, is used as an illustrative example. Probabilistic condition of the bridge is evaluated by the distribution of load rating factors obtained using multiple FE models. The numerical example shows that the proposed method can quantify uncertainty of measured data and subsequently evaluate efficiently the probabilistic condition of bridges.

• Journal of the Korean Data and Information Science Society
• /
• 제28권4호
• /
• pp.927-936
• /
• 2017

To analyze longitudinal count data, Poisson linear mixed models are commonly used. In the models the random effects covariance matrix explains both within-subject variation and serial correlation of repeated count outcomes. When the random effects covariance matrix is assumed to be misspecified, the estimates of covariates effects can be biased. Therefore, we propose reasonable and flexible structures of the covariance matrix using autoregressive and moving average Cholesky decomposition (ARMACD). The ARMACD factors the covariance matrix into generalized autoregressive parameters (GARPs), generalized moving average parameters (GMAPs) and innovation variances (IVs). Positive IVs guarantee the positive-definiteness of the covariance matrix. In this paper, we use the ARMACD to model the random effects covariance matrix in Poisson loglinear mixed models. We analyze epileptic seizure data using our proposed model.

• 스마트미디어저널
• /
• 제10권3호
• /
• pp.9-22
• /
• 2021

Basically, machine learning models use input data to produce results. Sometimes, the input data is too complicated for the models to learn useful patterns. Therefore, feature engineering is a crucial data preprocessing step for constructing a proper feature set to improve the performance of such models. One of the most efficient methods for automating feature engineering is the autoencoder, which transforms the data from its original space into a latent space. However certain factors, including the datasets, the machine learning models, and the number of dimensions of the latent space (denoted by k), should be carefully considered when using the autoencoder. In this study, we design a framework to compare two data preprocessing approaches: with and without autoencoder and to observe the impact of these factors on autoencoder. We then conduct experiments using autoencoders with classifiers on popular datasets. The empirical results provide a perspective regarding the best suited autoencoder for these factors.

• Communications for Statistical Applications and Methods
• /
• 제28권2호
• /
• pp.171-188
• /
• 2021

Due to an increased demand for longevity risk analysis, various stochastic models have been suggested to evaluate uncertainly in estimated life expectancy and the associated value of future annuity payments. Recently updated data allow us to analyze mortality for a longer historical period and extended age ranges. This study followed up previous case studies using up-to-date empirical data on Korean mortality and the recently developed R package StMoMo for stochastic mortality models analysis. The suitability of stochastic mortality models, focusing on retirement ages, was investigated with goodness-of-fit, validity of models, and ability of generating reasonable sets of simulation paths of future mortality. Comparisons were made across various types of models. Based on the selected models, the variability of important estimated measures associated with pension, annuity, and reverse mortgage were quantified using simulations.