• Journal of Intelligence and Information Systems
• /
• v.23 no.3
• /
• pp.139-153
• /
• 2017

Corporate bankruptcy can cause great losses not only to stakeholders but also to many related sectors in society. Through the economic crises, bankruptcy have increased and bankruptcy prediction models have become more and more important. Therefore, corporate bankruptcy has been regarded as one of the major topics of research in business management. Also, many studies in the industry are in progress and important. Previous studies attempted to utilize various methodologies to improve the bankruptcy prediction accuracy and to resolve the overfitting problem, such as Multivariate Discriminant Analysis (MDA), Generalized Linear Model (GLM). These methods are based on statistics. Recently, researchers have used machine learning methodologies such as Support Vector Machine (SVM), Artificial Neural Network (ANN). Furthermore, fuzzy theory and genetic algorithms were used. Because of this change, many of bankruptcy models are developed. Also, performance has been improved. In general, the company's financial and accounting information will change over time. Likewise, the market situation also changes, so there are many difficulties in predicting bankruptcy only with information at a certain point in time. However, even though traditional research has problems that don't take into account the time effect, dynamic model has not been studied much. When we ignore the time effect, we get the biased results. So the static model may not be suitable for predicting bankruptcy. Thus, using the dynamic model, there is a possibility that bankruptcy prediction model is improved. In this paper, we propose RNN (Recurrent Neural Network) which is one of the deep learning methodologies. The RNN learns time series data and the performance is known to be good. Prior to experiment, we selected non-financial firms listed on the KOSPI, KOSDAQ and KONEX markets from 2010 to 2016 for the estimation of the bankruptcy prediction model and the comparison of forecasting performance. In order to prevent a mistake of predicting bankruptcy by using the financial information already reflected in the deterioration of the financial condition of the company, the financial information was collected with a lag of two years, and the default period was defined from January to December of the year. Then we defined the bankruptcy. The bankruptcy we defined is the abolition of the listing due to sluggish earnings. We confirmed abolition of the list at KIND that is corporate stock information website. Then we selected variables at previous papers. The first set of variables are Z-score variables. These variables have become traditional variables in predicting bankruptcy. The second set of variables are dynamic variable set. Finally we selected 240 normal companies and 226 bankrupt companies at the first variable set. Likewise, we selected 229 normal companies and 226 bankrupt companies at the second variable set. We created a model that reflects dynamic changes in time-series financial data and by comparing the suggested model with the analysis of existing bankruptcy predictive models, we found that the suggested model could help to improve the accuracy of bankruptcy predictions. We used financial data in KIS Value (Financial database) and selected Multivariate Discriminant Analysis (MDA), Generalized Linear Model called logistic regression (GLM), Support Vector Machine (SVM), Artificial Neural Network (ANN) model as benchmark. The result of the experiment proved that RNN's performance was better than comparative model. The accuracy of RNN was high in both sets of variables and the Area Under the Curve (AUC) value was also high. Also when we saw the hit-ratio table, the ratio of RNNs that predicted a poor company to be bankrupt was higher than that of other comparative models. However the limitation of this paper is that an overfitting problem occurs during RNN learning. But we expect to be able to solve the overfitting problem by selecting more learning data and appropriate variables. From these result, it is expected that this research will contribute to the development of a bankruptcy prediction by proposing a new dynamic model.

• Structural Engineering and Mechanics
• /
• v.7 no.4
• /
• pp.413-432
• /
• 1999

A macro-element model is developed to account for shear deformation and bond slip of reinforcement bars in the beam-column joint region of reinforced concrete structures. The joint region is idealized by two springs in series, one representing shear deformation and the other representing bond slip. The softened truss model theory is adopted to establish the shear force-shear deformation relationship and to determine the shear capacity of the joint. A detailed model for the bond slip of the reinforcing bars at the beam-column interface is presented. The proposed macro-element model of the joint is validated using available experimental data on beam-column connections representing exterior joints in ductile and nonductile frames.

• Structural Engineering and Mechanics
• /
• v.11 no.6
• /
• pp.651-662
• /
• 2001

In order to investigate the seismic behavior and seismic design methods in the transverse direction of a shield tunnel, a series of model shaking table tests and a two-dimensional finite element dynamic analysis on the tests are carried out. Two kinds of static analytical methods based on ground-tunnel composite finite element model and beam-spring element model are proposed, and the validity of the static analyses is verified by model shaking table tests. The investigation concerns the dynamic response behavior of a tunnel and the ground, the interaction between the tunnel and ground, and an evaluation of different seismic design methods. Results of the investigation indicate that the shield tunnel follows the surrounding ground in displacement and dynamic characteristics in the transverse direction; also, the static analytical methods proposed by the authors can be used directly as the seismic design methods in the transverse direction of a shield tunnel.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.1360-1364
• /
• 2004

This paper propose an extended model for role-permission assignment based on locations called "Enhanced Role-Based Access Control (ERBAC03)". The proposed model is built upon the well-known RBAC model. Assigning permissions to role is considered too complex activity to accomplish directly. Instead we advocate breaking down this process into a number of steps. The concept of jobs and tasks is specifically introduced to facilitate role-permission assignment into a series of smaller steps. This model is suitable for any large organization that has many branches. Each branch consists of many users who work in difference roles. An administration tool has been developed to assist administrators with the administration of separation of duty requirements. It demonstrates how the specification of static requirements can be done based on "conflicting entities" paradigm. Static separation of duty requirements must be enforced in the administration environment. Finally, we illustrate how the ERBAC03 prototype is used to administer the separation of duty requirements.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 1995.04a
• /
• pp.53-56
• /
• 1995

Despite the recent development of structural analysis programs for the CRCP pavements over Westergaard's equations and finite element techniques, the Winkler foundations which are modelled by series of vertical springs at the nodes are generally used for the computer modelling of subbases under the concrete slab. Herewith, two parameter of soil foundation model is adopted as the most convenient mathematical model to enable deflections outside the loaded area to be effected and to upgrade the Winkler foundations. This paper highlights the derivations of finite element method for the two-parameter soil foundation model in the concrete pavements.

• Geomechanics and Engineering
• /
• v.8 no.4
• /
• pp.595-613
• /
• 2015

An experimental investigation into the uplift capacity of horizontal square plate anchors in sand with and without geogrid reinforcement is reported. The parameters investigated are the effect of the depth of the single layer of geogrid, vertical spacing of geogrid layers, number of geogrid layers, length of geogrid layers, the effects of embedment depth, and relative density of sand. A series of three dimensional finite element analyses model was established and confirmed to be effective in capturing the behaviour of plate anchor-reinforced sand by comparing its predictions with experimental results. The results showed that the geogrid reinforcement had a considerable effect on the uplift capacity of horizontal square plate anchors in sand. The improvement in uplift capacity was found to be strongly dependent on the embedment depth and relative density of sand. A satisfactory agreement between the experimental and numerical results on general trend of behaviour and optimum geometry of reinforcement placement is observed. Based on the model test results and the finite element analyses, optimum values of the geogrid parameters for maximum reinforcing effect are discussed and suggested.

• Communications for Statistical Applications and Methods
• /
• v.22 no.6
• /
• pp.615-624
• /
• 2015

We consider the estimation of seasonal cointegration in the presence of conditional heteroskedasticity (CH) using a feasible generalized least squares method. We capture cointegrating relationships and time-varying volatility for long-run and short-run dynamics in the same model. This procedure can be easily implemented using common methods such as ordinary least squares and generalized least squares. The maximum likelihood (ML) estimation method is computationally difficult and may not be feasible for larger models. The simulation results indicate that the proposed method is superior to the ML method when CH exists. In order to illustrate the proposed method, an empirical example is presented to model a seasonally cointegrated times series under CH.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2001.10a
• /
• pp.181-185
• /
• 2001

Three-dimensional finite-element-based numerical model of turbulent flow, heat transfer, macroscopic solidification and inclusion trajectory in a continuos steel slab caster was developed Turbulence was incorporated using the Improved Low-Re turbulence model with positive preserving approach. The mushy region was modeled as the porous media with average effective viscosity. A series of simulations was carried out to investigate the effects of the casting speed, the slab size, the delivered superheat the immersion depth of the SEN on the transport phenomena. In the absence of any known experimental data related to velocity profiles, the numerical predictions of the solidified profile on a caster was compared with breakouts data and a good agreement was found.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2001.10a
• /
• pp.125-128
• /
• 2001

A finite element-based, integrated process model is presented for coupled analysis of the thermal and metallurgical behavior of the strip occurring on the run-out-table in hot strip rolling. The validity of the proposed model is examined through comparison with measurements. The models capability of revealing the effect of cooling pattern on strip temperatures and the optimal cooling pattern are demonstrated through a series of process simulation. In order to improve strip shape and control temperature history of thickness direction for strip during ROT cooling.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2001.10a
• /
• pp.129-133
• /
• 2001

A finite element-based, integrated process model is presented for a three dimensional, coupled analysis of the thermo-mechanical behavior of the strip and work roll in the continuous hot strip rolling. The validity of the proposed model is examined through comparison with measurements. The effect of Edge-Heater on the finishing delivery temperatures is examined by using the present model. The models capability of revealing the effect of diverse process parameters is demonstrated through a series of process simulation.