• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.3
• /
• pp.107-114
• /
• 2021

Recently artificial intelligence technology has been introduced in various fields and various machine learning models have been operated in various frameworks as academic interest has increased. However, these frameworks have different data formats, which lack interoperability, and to overcome this, the open neural network exchange format, ONNX, has been proposed. In this paper we describe how to transform multiple machine learning models to ONNX, and propose algorithms and inference systems that can determine machine learning techniques in an integrated ONNX format. Furthermore we compare the inference results of the models before and after the ONNX transformation, showing that there is no loss or performance degradation of the learning results between the ONNX transformation.

• The Korean Journal of Applied Statistics
• /
• v.21 no.3
• /
• pp.537-546
• /
• 2008

One of main goals of time series analysis is to estimate prediction of future values. In this paper, we investigate the bias problem when the transformation and back- transformation approach is applied in ARMA models and introduce a modified smearing estimation to reduce the bias. An empirical study on the returns of KOSDAQ index via Yeo-Johnson transformation was executed to compare the performance of existing methods and proposed methods and showed that proposed approaches provide a bias-reduced estimation of the prediction value.

• Journal of Korean Society of Forest Science
• /
• v.88 no.1
• /
• pp.47-54
• /
• 1999

To estimate optimal tree diameter and height growth function in Pinus thunbergii stands with site index of 12 class, quoted from two linear models of linear transformation(1) and linear transformation (2) and four non-linear models of exponential, Gompertz, Chapman-Richards, and Weibull etc.. Analyzed correlation among the estimated tree diameter and height by these function models, and observed diameter and height growth were compared. In the results of tree diameter and height growth estimation by stand age, non-linear models showed better appropriation than linear model and Chapman-Richards model was most fitted for tree height growth but few, if any, differences among their nonlinear models. Therefore, it is consider to be much more study about non-linear model to estimate tree diameter and height growth in the actual stands hereafter.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.10
• /
• pp.17-25
• /
• 2023

In this paper, we propose a lightweight deep learning architecture tailored for efficient text detection in fashion design sketch images. Given the increasing prominence of Digital Transformation in the fashion industry, there is a growing emphasis on harnessing digital tools for creating fashion design sketches. As digitization becomes more pervasive in the fashion design process, the initial stages of text detection and recognition take on pivotal roles. In this study, a lightweight network was designed by building upon existing text detection deep learning models, taking into consideration the unique characteristics of apparel design drawings. Additionally, a separately collected dataset of apparel design drawings was added to train the deep learning model. Experimental results underscore the superior performance of our proposed deep learning model, outperforming existing text detection models by approximately 20% when applied to fashion design sketch images. As a result, this paper is expected to contribute to the Digital Transformation in the field of clothing design by means of research on optimizing deep learning models and detecting specialized text information.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.23 no.2
• /
• pp.103-129
• /
• 1998

Corporate level decision making with multiple decision makers in a consistent way is essential in Decision Support System. However, since the decision makers have different interests and knowledge, the models used by them are also different in their level of abstraction. This makes decision makers waste a lot of efforts for an integrated decision making. The purpose of this paper is to propose an integration mechanism so that collaborative decision making models may be used synthetically in multi-abstraction level. Models are classified as multimedia model, mathematical model, qualitative model, causal ＆ directional model, causal model, directional model and relationship model according to the level of abstraction. The proposed integration mechanism consists of model interpretation phase. model transformation phase, and model integration phase. Specifically, the model transformation Phase is divided into (1) model tightening mode which gather information to make a model transformed into upper level model, and (2) model relaxing mode which makes lower level model. In the model integration phase, models of same level are to be integrated schematically. An illustrative M＆A-decision example is given to show the possibility of the methodology.

• The Korean Journal of Applied Statistics
• /
• v.33 no.6
• /
• pp.713-722
• /
• 2020

Contrasted with the standard symmetric GARCH models, we consider a broad class of threshold-asymmetric models to analyse financial time series exhibiting asymmetric volatility. By further introducing power transformations, we add more flexibilities to the asymmetric class, thereby leading to power transformed and asymmetric volatility models. In particular, the paper is concerned with the nonstationary volatilities in which conditions for integrated volatility and explosive volatility are separately discussed. Dow Jones Industrial Average is analysed for illustration.

• Journal of the Korean Ceramic Society
• /
• v.49 no.1
• /
• pp.1-10
• /
• 2012

Super-plastic deformation that originates from transformation plasticity has favorable aspects for steels with improved strength and ductility. However, it also causes undesirable deformation of products or specimens, leading to their degradation. This article reviews recent investigations of transformation plasticity. A combination of newly suggested models, numerical analyses, and novel experiments has attempted to reveal the mechanism. Since the nature of the transformation plasticity is still unclear, there are significant challenges still to be solved. Fundamental understanding of transformation plasticity will be essential for the development of advanced steels.

• Journal of the Korea Society for Simulation
• /
• v.1 no.1
• /
• pp.77-86
• /
• 1992

In this paper, we study validation test methods of DEVSA(Descrete Event system Specification) models using SPN(Stochastic Petri Net) models. We discuss conventional validation test methods, by which DEVS models can be transformed to SPN models, by reviewing the features of DEVS model. Based on the model transformation method, we define a new homogeneous function for validation test and suggest a new validation test method of DEVS models using the property of SPN models and the new homogeneous function.

• Structural Engineering and Mechanics
• /
• v.46 no.6
• /
• pp.755-773
• /
• 2013

One of the main requirements of the seismic design codes must be its easy application by structural engineers. The use of practically-applicable models or simplified models as single-degree-of-freedom (SDOF) systems is a good alternative to achieve this condition. In this study, deterministic and probabilistic response transformation factors are obtained to evaluate the response in terms of maximum ductility and maximum interstory drifts of multi-degree-of-freedom (MDOF) systems based on the response of equivalent SDOF systems. For this aim, five steel frames designed with the Mexican City Building Code (MCBC) as well as their corresponding equivalent SDOF systems (which represent the characteristics of the frames) are analyzed. Both structural systems are subjected to ground motions records. For the MDOF and the simplified systems, incremental dynamic analyses IDAs are developed in first place, then, structural demand hazard curves are obtained. The ratio between the IDAs curves corresponding to the MDOF systems and the curves corresponding to the simplified models are used to obtain deterministic response transformation factors. On the other hand, demand hazard curves are used to calculate probabilistic response transformation factors. It was found that both approaches give place to similar results.

• The KIPS Transactions:PartD
• /
• v.18D no.6
• /
• pp.461-480
• /
• 2011

Component-Based Development (CBD) focuses on developing reusable components and assembling them into target applications. Service-Oriented Computing (SOC) is a relatively new paradigm where reusable services are developed, published, and composed into target service applications. Due to the popularity of CBD, many organizations already have CBD models for their applications. Hence, it is feasible and appealing to develop SOC applications economically by transforming existing CBD models, rather than developing SOC applications from the scratch. Our research is to develop a methodology for transforming existing CBD design into Service-Oriented Architecture (SOA) design. In this paper, we first compare the two paradigms and their key elements for the design; CBD design and SOA design. Then, we define mapping relationships of the transformation. For each transformation mapping, we present step-wise instructions. The benefit of applying this transformation is to increase reusability of existing design assets, to maintain the consistency between existing CBD and SOC models, and reduced cost for deploying service-oriented applications.