• Journal of Intelligence and Information Systems
• /
• v.20 no.4
• /
• pp.121-139
• /
• 2014

Predicting corporate failure has been an important topic in accounting and finance. The costs associated with bankruptcy are high, so the accuracy of bankruptcy prediction is greatly important for financial institutions. Lots of researchers have dealt with the topic associated with bankruptcy prediction in the past three decades. The current research attempts to use ensemble models for improving the performance of bankruptcy prediction. Ensemble classification is to combine individually trained classifiers in order to gain more accurate prediction than individual models. Ensemble techniques are shown to be very useful for improving the generalization ability of the classifier. Bagging is the most commonly used methods for constructing ensemble classifiers. In bagging, the different training data subsets are randomly drawn with replacement from the original training dataset. Base classifiers are trained on the different bootstrap samples. Instance selection is to select critical instances while deleting and removing irrelevant and harmful instances from the original set. Instance selection and bagging are quite well known in data mining. However, few studies have dealt with the integration of instance selection and bagging. This study proposes an improved bagging ensemble based on instance selection using genetic algorithms (GA) for improving the performance of SVM. GA is an efficient optimization procedure based on the theory of natural selection and evolution. GA uses the idea of survival of the fittest by progressively accepting better solutions to the problems. GA searches by maintaining a population of solutions from which better solutions are created rather than making incremental changes to a single solution to the problem. The initial solution population is generated randomly and evolves into the next generation by genetic operators such as selection, crossover and mutation. The solutions coded by strings are evaluated by the fitness function. The proposed model consists of two phases: GA based Instance Selection and Instance based Bagging. In the first phase, GA is used to select optimal instance subset that is used as input data of bagging model. In this study, the chromosome is encoded as a form of binary string for the instance subset. In this phase, the population size was set to 100 while maximum number of generations was set to 150. We set the crossover rate and mutation rate to 0.7 and 0.1 respectively. We used the prediction accuracy of model as the fitness function of GA. SVM model is trained on training data set using the selected instance subset. The prediction accuracy of SVM model over test data set is used as fitness value in order to avoid overfitting. In the second phase, we used the optimal instance subset selected in the first phase as input data of bagging model. We used SVM model as base classifier for bagging ensemble. The majority voting scheme was used as a combining method in this study. This study applies the proposed model to the bankruptcy prediction problem using a real data set from Korean companies. The research data used in this study contains 1832 externally non-audited firms which filed for bankruptcy (916 cases) and non-bankruptcy (916 cases). Financial ratios categorized as stability, profitability, growth, activity and cash flow were investigated through literature review and basic statistical methods and we selected 8 financial ratios as the final input variables. We separated the whole data into three subsets as training, test and validation data set. In this study, we compared the proposed model with several comparative models including the simple individual SVM model, the simple bagging model and the instance selection based SVM model. The McNemar tests were used to examine whether the proposed model significantly outperforms the other models. The experimental results show that the proposed model outperforms the other models.

• Journal of Practical Engineering Education
• /
• v.15 no.1
• /
• pp.119-126
• /
• 2023

Today, in order to improve fuel efficiency and dynamic behavior of automobiles, an era of light weight and simplification of automobile parts is being formed. In order to simplify and design and manufacture the shape of the product, various components are integrated. For example, in order to commercialize three products into one product, product processing is occurring to a very narrow area. In the case of existing parts, precision die casting or casting production is used for processing convenience, and the multi-piece method requires a lot of processes and reduces the precision and strength of the parts. It is very advantageous to manufacture integrally to simplify the processing air and secure the strength of the parts, but if a deep and narrow pocket part needs to be processed, it cannot be processed with the equipment's own spindle. To solve a problem, research on cutting processing is being actively conducted, and multi-axis composite processing technology not only solves this problem. It has many advantages, such as being able to cut into composite shapes that have been difficult to flexibly cut through various processes with one machine tool so far. However, the reality is that expensive equipment increases manufacturing costs and lacks engineers who can operate the machine. In the five-axis cutting processing machine, when producing products with deep and narrow sections, the cycle time increases in product production due to the indirectness of tools, and many problems occur in processing. Therefore, dedicated machine tools and multi-axis composite machines should be used. Alternatively, an angle spindle may be used as a special tool capable of multi-axis composite machining of five or more axes in a three-axis machining center. Various and continuous studies are needed in areas such as processing vibration absorption, low heat generation and operational stability, excellent dimensional stability, and strength securing by using the angle spindle.

• Journal of Intelligence and Information Systems
• /
• v.18 no.2
• /
• pp.29-45
• /
• 2012

Bond rating is regarded as an important event for measuring financial risk of companies and for determining the investment returns of investors. As a result, it has been a popular research topic for researchers to predict companies' credit ratings by applying statistical and machine learning techniques. The statistical techniques, including multiple regression, multiple discriminant analysis (MDA), logistic models (LOGIT), and probit analysis, have been traditionally used in bond rating. However, one major drawback is that it should be based on strict assumptions. Such strict assumptions include linearity, normality, independence among predictor variables and pre-existing functional forms relating the criterion variablesand the predictor variables. Those strict assumptions of traditional statistics have limited their application to the real world. Machine learning techniques also used in bond rating prediction models include decision trees (DT), neural networks (NN), and Support Vector Machine (SVM). Especially, SVM is recognized as a new and promising classification and regression analysis method. SVM learns a separating hyperplane that can maximize the margin between two categories. SVM is simple enough to be analyzed mathematical, and leads to high performance in practical applications. SVM implements the structuralrisk minimization principle and searches to minimize an upper bound of the generalization error. In addition, the solution of SVM may be a global optimum and thus, overfitting is unlikely to occur with SVM. In addition, SVM does not require too many data sample for training since it builds prediction models by only using some representative sample near the boundaries called support vectors. A number of experimental researches have indicated that SVM has been successfully applied in a variety of pattern recognition fields. However, there are three major drawbacks that can be potential causes for degrading SVM's performance. First, SVM is originally proposed for solving binary-class classification problems. Methods for combining SVMs for multi-class classification such as One-Against-One, One-Against-All have been proposed, but they do not improve the performance in multi-class classification problem as much as SVM for binary-class classification. Second, approximation algorithms (e.g. decomposition methods, sequential minimal optimization algorithm) could be used for effective multi-class computation to reduce computation time, but it could deteriorate classification performance. Third, the difficulty in multi-class prediction problems is in data imbalance problem that can occur when the number of instances in one class greatly outnumbers the number of instances in the other class. Such data sets often cause a default classifier to be built due to skewed boundary and thus the reduction in the classification accuracy of such a classifier. SVM ensemble learning is one of machine learning methods to cope with the above drawbacks. Ensemble learning is a method for improving the performance of classification and prediction algorithms. AdaBoost is one of the widely used ensemble learning techniques. It constructs a composite classifier by sequentially training classifiers while increasing weight on the misclassified observations through iterations. The observations that are incorrectly predicted by previous classifiers are chosen more often than examples that are correctly predicted. Thus Boosting attempts to produce new classifiers that are better able to predict examples for which the current ensemble's performance is poor. In this way, it can reinforce the training of the misclassified observations of the minority class. This paper proposes a multiclass Geometric Mean-based Boosting (MGM-Boost) to resolve multiclass prediction problem. Since MGM-Boost introduces the notion of geometric mean into AdaBoost, it can perform learning process considering the geometric mean-based accuracy and errors of multiclass. This study applies MGM-Boost to the real-world bond rating case for Korean companies to examine the feasibility of MGM-Boost. 10-fold cross validations for threetimes with different random seeds are performed in order to ensure that the comparison among three different classifiers does not happen by chance. For each of 10-fold cross validation, the entire data set is first partitioned into tenequal-sized sets, and then each set is in turn used as the test set while the classifier trains on the other nine sets. That is, cross-validated folds have been tested independently of each algorithm. Through these steps, we have obtained the results for classifiers on each of the 30 experiments. In the comparison of arithmetic mean-based prediction accuracy between individual classifiers, MGM-Boost (52.95%) shows higher prediction accuracy than both AdaBoost (51.69%) and SVM (49.47%). MGM-Boost (28.12%) also shows the higher prediction accuracy than AdaBoost (24.65%) and SVM (15.42%)in terms of geometric mean-based prediction accuracy. T-test is used to examine whether the performance of each classifiers for 30 folds is significantly different. The results indicate that performance of MGM-Boost is significantly different from AdaBoost and SVM classifiers at 1% level. These results mean that MGM-Boost can provide robust and stable solutions to multi-classproblems such as bond rating.

• Journal of Intelligence and Information Systems
• /
• v.26 no.4
• /
• pp.27-65
• /
• 2020

Many companies on information and communication technology make public their own developed AI technology, for example, Google's TensorFlow, Facebook's PyTorch, Microsoft's CNTK. By releasing deep learning open source software to the public, the relationship with the developer community and the artificial intelligence (AI) ecosystem can be strengthened, and users can perform experiment, implementation and improvement of it. Accordingly, the field of machine learning is growing rapidly, and developers are using and reproducing various learning algorithms in each field. Although various analysis of open source software has been made, there is a lack of studies to help develop or use deep learning open source software in the industry. This study thus attempts to derive a strategy for adopting the framework through case studies of a deep learning open source framework. Based on the technology-organization-environment (TOE) framework and literature review related to the adoption of open source software, we employed the case study framework that includes technological factors as perceived relative advantage, perceived compatibility, perceived complexity, and perceived trialability, organizational factors as management support and knowledge & expertise, and environmental factors as availability of technology skills and services, and platform long term viability. We conducted a case study analysis of three companies' adoption cases (two cases of success and one case of failure) and revealed that seven out of eight TOE factors and several factors regarding company, team and resource are significant for the adoption of deep learning open source framework. By organizing the case study analysis results, we provided five important success factors for adopting deep learning framework: the knowledge and expertise of developers in the team, hardware (GPU) environment, data enterprise cooperation system, deep learning framework platform, deep learning framework work tool service. In order for an organization to successfully adopt a deep learning open source framework, at the stage of using the framework, first, the hardware (GPU) environment for AI R&D group must support the knowledge and expertise of the developers in the team. Second, it is necessary to support the use of deep learning frameworks by research developers through collecting and managing data inside and outside the company with a data enterprise cooperation system. Third, deep learning research expertise must be supplemented through cooperation with researchers from academic institutions such as universities and research institutes. Satisfying three procedures in the stage of using the deep learning framework, companies will increase the number of deep learning research developers, the ability to use the deep learning framework, and the support of GPU resource. In the proliferation stage of the deep learning framework, fourth, a company makes the deep learning framework platform that improves the research efficiency and effectiveness of the developers, for example, the optimization of the hardware (GPU) environment automatically. Fifth, the deep learning framework tool service team complements the developers' expertise through sharing the information of the external deep learning open source framework community to the in-house community and activating developer retraining and seminars. To implement the identified five success factors, a step-by-step enterprise procedure for adoption of the deep learning framework was proposed: defining the project problem, confirming whether the deep learning methodology is the right method, confirming whether the deep learning framework is the right tool, using the deep learning framework by the enterprise, spreading the framework of the enterprise. The first three steps (i.e. defining the project problem, confirming whether the deep learning methodology is the right method, and confirming whether the deep learning framework is the right tool) are pre-considerations to adopt a deep learning open source framework. After the three pre-considerations steps are clear, next two steps (i.e. using the deep learning framework by the enterprise and spreading the framework of the enterprise) can be processed. In the fourth step, the knowledge and expertise of developers in the team are important in addition to hardware (GPU) environment and data enterprise cooperation system. In final step, five important factors are realized for a successful adoption of the deep learning open source framework. This study provides strategic implications for companies adopting or using deep learning framework according to the needs of each industry and business.

• Journal of Korean Society of Transportation
• /
• v.18 no.1
• /
• pp.75-85
• /
• 2000

In existing models in optimization, the crisp data improve has been used in the objective or constraints to derive the optimal solution, Besides, the subjective environments are eliminated because the complex and uncertain circumstances were regarded as Probable ambiguity, In other words those optimal solutions in the existing models could be the complete satisfactory solutions to the objective functions in the Process of application for industrial engineering methods to minimize risks of decision-making. As a result of those, decision-makers in location Problems couldn't face appropriately with the variation of demand as well as other variables and couldn't Provide the chance of wide selection because of the insufficient information. So under the circumstance. it has been to develop the model for the location and size decision problems of logistics facility in the use of the fuzzy theory in the intention of making the most reasonable decision in the Point of subjective view under ambiguous circumstances, in the foundation of the existing decision-making problems which must satisfy the constraints to optimize the objective function in strictly given conditions in this study. Introducing the Process used in this study after the establishment of a general mixed integer Programming(MIP) model based upon the result of existing studies to decide the location and size simultaneously, a fuzzy mixed integer Programming(FMIP) model has been developed in the use of fuzzy theory. And the general linear Programming software, LINDO 6.01 has been used to simulate, to evaluate the developed model with the examples and to judge of the appropriateness and adaptability of the model(FMIP) in the real world.

• Journal of the Korean Crystal Growth and Crystal Technology
• /
• v.8 no.4
• /
• pp.599-604
• /
• 1998

We have investigated Pt and $RuO_2$ as a bottom electrode for ferroelectric capacitor applications. The bottom electrodes were prepared by using an RF magnetron sputtering method. Some of the investigated parameters were a substrate temperature, gas flow rate, RF power for the film growth, and post annealing effect. The substrate temperature strongly influenced the surface morphology and resistivity of the bottom electrodes as well as the film crystallographic structure. XRD results on Pt films showed a mixed phase of (111) and (200) peak for the substrate temperature ranged from RT to $200^{\circ}C$, and a preferred (111) orientation for $300^{\circ}C$. From the XRD and AFM results, we recommend the substrate temperature of $300^{\circ}C$ and RF power 80W for the Pt bottom electrode growth. With the variation of an oxygen partial pressure from 0 to 50%, we learned that only Ru metal was grown with 0~5% of $O_2$ gas, mixed phase of Ru and $RuO_2$ for $O_ 2$ partial pressure between 10~40%, and a pure $RuO_2$ phase with $O_2$ partial pressure of 50%. This result indicates that a double layer of $RuO_2/Ru$ can be grown in a process with the modulation of gas flow rate. Double layer structure is expected to reduce the fatigue problem while keeping a low electrical resistivity. As post anneal temperature was increased from RT to $700^{\circ}C$, the resistivity of Pt and $RuO_2$ was decreased linearly. This paper presents the optimized process conditions of the bottom electrodes for memory device applications.

• Radiation Oncology Journal
• /
• v.24 no.4
• /
• pp.300-308
• /
• 2006

$\underline{Purpose}$: To develop a wireless CCTV system in semi-beam's eye view (BEV) to monitor daily patient setup in radiation therapy. $\underline{Materials\;and\;Methods}$: In order to get patient images in semi-BEV, CCTV cameras are installed in a custom-made acrylic applicator below the treatment head of a linear accelerator. The images from the cameras are transmitted via radio frequency signal (${\sim}2.4\;GHz$ and 10 mW RF output). An expected problem with this system is radio frequency interference, which is solved utilizing RF shielding with Cu foils and median filtering software. The images are analyzed by our custom-made software. In the software, three anatomical landmarks in the patient surface are indicated by a user, then automatically the 3 dimensional structures are obtained and registered by utilizing a localization procedure consisting mainly of stereo matching algorithm and Gauss-Newton optimization. This algorithm is applied to phantom images to investigate the setup accuracy. Respiratory gating system is also researched with real-time image processing. A line-laser marker projected on a patient's surface is extracted by binary image processing and the breath pattern is calculated and displayed in real-time. $\underline{Results}$: More than 80% of the camera noises from the linear accelerator are eliminated by wrapping the camera with copper foils. The accuracy of the localization procedure is found to be on the order of $1.5{\pm}0.7\;mm$ with a point phantom and sub-millimeters and degrees with a custom-made head/neck phantom. With line-laser marker, real-time respiratory monitoring is possible in the delay time of ${\sim}0.17\;sec$. $\underline{Conclusion}$: The wireless CCTV camera system is the novel tool which can monitor daily patient setups. The feasibility of respiratory gating system with the wireless CCTV is hopeful.

• The KIPS Transactions:PartD
• /
• v.12D no.3 s.99
• /
• pp.417-428
• /
• 2005

Today, IT organizations perform projects with vision related to marketing and financial profit. The objective of realizing the vision is to improve the project performing ability in terms of QCD. Organizations have made a lot of efforts to achieve this objective through process improvement. Large companies such as IBM, Ford, and GE have made over $80\%$ of success through business process re-engineering using information technology instead of business improvement effect by computers. It is important to collect, analyze and manage the data on performed projects to achieve the objective, but quantitative measurement is difficult as software is invisible and the effect and efficiency caused by process change are not visibly identified. Therefore, it is not easy to extract the strategy of improvement. This paper measures and analyzes the project performance, focusing on organizations' external effectiveness and internal efficiency (Qualify, Delivery, Cycle time, and Waste). Based on the measured project performance scores, an OT (Opportunity Tree) model was designed for optimizing the project performance. The process of design is as follows. First, meta data are derived from projects and analyzed by quantitative GQM(Goal-Question-Metric) questionnaire. Then, the project performance model is designed with the data obtained from the quantitative GQM questionnaire and organization's performance score for each area is calculated. The value is revised by integrating the measured scores by area vision weights from all stakeholders (CEO, middle-class managers, developer, investor, and custom). Through this, routes for improvement are presented and an optimized improvement method is suggested. Existing methods to improve software process have been highly effective in division of processes' but somewhat unsatisfactory in structural function to develop and systemically manage strategies by applying the processes to Projects. The proposed OT model provides a solution to this problem. The OT model is useful to provide an optimal improvement method in line with organization's goals and can reduce risks which may occur in the course of improving process if it is applied with proposed methods. In addition, satisfaction about the improvement strategy can be improved by obtaining input about vision weight from all stakeholders through the qualitative questionnaire and by reflecting it to the calculation. The OT is also useful to optimize the expansion of market and financial performance by controlling the ability of Quality, Delivery, Cycle time, and Waste.

• KDI Journal of Economic Policy
• /
• v.12 no.1
• /
• pp.127-148
• /
• 1990

A Milgrom-Roberts style signalling model of limit pricing is developed to analyze the possibility and the scope of limit pricing in general, noncooperative oligopolies. The model contains multiple incumbent firms facing a potential entrant and assumes an information asymmetry between incombents and the potential entrant about the market demand. There are two periods in the model. In period 1, n incumbent firms simultaneously and noncooperatively choose quantities. At the end of period 1, the potential entrant observes the market price and makes an entry decision. In period 2, depending on the entry decision of the entrant, n' or (n+1) firms choose quantities again before the game terminates. Since the choice of incumbent firms in period 1 depends on their information about demand, the market price in period 1 conveys information about the market demand. Thus, there is a systematic link between the market price and the profitability of entry. Using Bayes-Nash equilibrium as the solution concept, we find that there exist some demand conditions under which incumbent firms will limit price. In symmetric equilibria, incumbent firms each produce an output that is greater than the Cournot output and induce a price that is below the Cournot price. In doing so, each incumbent firm refrains from maximizing short-run profit and supplies a public good that is entry deterrence. The reason that entry is deterred by such a reduced price is that it conveys information about the demand of the industry that is unfavorable to the entrant. This establishes the possibility of limit pricing by noncooperative oligopolists in a setting that is fully rational, and also generalizes the result of Milgrom and Roberts to general oligopolies, confirming Bain's intuition. Limit pricing by incumbents explained above can be interpreted as a form of credible collusion in which each firm voluntarily deviates from myopic optimization in order to deter entry using their superior information. This type of implicit collusion differs from Folk-theorem type collusions in many ways and suggests that a collusion can be a credible one even in finite games as long as there is information asymmetry. Another important result is that as the number of incumbent firms approaches infinity, or as the industry approaches a competitive one, the probability that limit pricing occurs converges to zero and the probability of entry converges to that under complete information. This limit result confirms the intuition that as the number of agents sharing the same private information increases, the value of the private information decreases, and the probability that the information gets revealed increases. This limit result also supports the conventional belief that there is no entry problem in a competitive market. Considering the fact that limit pricing is generally believed to occur at an early stage of an industry and the fact that many industries in Korea are oligopolies in their infant stages, the theoretical results of this paper suggest that we should pay attention to the possibility of implicit collusion by incumbent firms aimed at deterring new entry using superior information. The long-term loss to the Korean economy from limit pricing can be very large if the industry in question is a part of the world market and the domestic potential entrant whose entry is deterred could .have developed into a competitor in the world market. In this case, the long-term loss to the Korean economy should include the lost opportunity in the world market in addition to the domestic long-run welfare loss.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.42 no.7
• /
• pp.1125-1132
• /
• 2013

In the kimchi manufacturing process, the starter is cultured on a large-scale and needs to be supplied at a low price to kimchi factories. However, current high costs associated with the culture of lactic acid bacteria for the starter, have led to rising kimchi prices. To solve this problem, the development of a new medium for culturing lactic acid bacteria was studied. The base materials of a this novel medium consisted of Chinese cabbage extract, a carbon source, a nitrogen source, and inorganic salts. The optimal composition of this medium was determined to be 30% Chinese cabbage extract, 2% maltose, 0.25% yeast extract, and $2{\times}$ salt stock (2% sodium acetate trihydrate, 0.8% disodium hydrogen phosphate, 0.8% sodium citrate, 0.8% ammonium sulfate, 0.04% magnesium sulfate, 0.02% manganese sulfate). The newly developed medium was named MFL (medium for lactic acid bacteria). After culture for 24 hr at $30^{\circ}C$, the CFU/mL of Leuconostoc (Leuc.) citreum GR1 in MRS and MFL was $3.41{\times}10^9$ and $7.49{\times}10^9$, respectively. The number of cells in the MFL medium was 2.2 times higher than their number in the MRS media. In a scale-up process using this optimized medium, the fermentation conditions for Leuc. citreum GR1 were tested in a 2 L working volume using a 5 L jar fermentor at $30^{\circ}C$. At an impeller speed of 50 rpm (without pH control), the viable cell count was $8.60{\times}10^9$ CFU/mL. From studies on pH-stat control fermentation, the optimal pH and regulating agent was determined to be 6.8 and NaOH, respectively. At an impeller speed of 50 rpm with pH control, the viable cell count was $11.42{\times}10^9(1.14{\times}10^{10})$ CFU/mL after cultivation for 20 hr - a value was 3.34 times higher than that obtained using the MRS media in biomass production. This MFL media is expected to have economic advantages for the cultivation of Leuc. citreum GR1 as a starter for kimchi production.