• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.8
• /
• pp.333-340
• /
• 2023

Efficient prediction of corporate bankruptcy is an important part of making appropriate lending decisions for financial institutions and reducing loan default rates. In many studies, classification models using artificial intelligence technology have been used. In the financial industry, even if the performance of the new predictive models is excellent, it should be accompanied by an intuitive explanation of the basis on which the result was determined. Recently, the US, EU, and South Korea have commonly presented the right to request explanations of algorithms, so transparency in the use of AI in the financial sector must be secured. In this paper, an artificial intelligence-based interpretable classification prediction model was proposed using corporate bankruptcy data that was open to the outside world. First, data preprocessing, 5-fold cross-validation, etc. were performed, and classification performance was compared through optimization of 10 supervised learning classification models such as logistic regression, SVM, XGBoost, and LightGBM. As a result, LightGBM was confirmed as the best performance model, and SHAP, an explainable artificial intelligence technique, was applied to provide a post-explanation of the bankruptcy prediction process.

• Proceedings of the KSME Conference
• /
• 2003.11a
• /
• pp.1430-1435
• /
• 2003

Satellite upper platform is optimized by response surface method which has non-gradient, semi-glogal, discrete and fast convergency characteristics. Sampling points are extracted by design of experiments using Central Composite Method and Factorial Design. Also response surface is generated by the various regression functions. Structure analysis is execuated with regard for static and dynamic environment in launching stage. As a result response surface method is superior to other optimization method with respect to optimum value and cost of computation time. Also a confidence is varified in the various regression models.

• Proceedings of the Korean Statistical Society Conference
• /
• 2003.05a
• /
• pp.19-24
• /
• 2003

신경망은 적용 다양성과 제약조건의 최소성, 강력한 예측성, 범용성, 근사성 등 많은 장점을 지니고 있으나 초기 가중치의 할당에 따라 모델 생성의 Performance와 예측의 결과가 달라지게 되는 단점을 지니고 있다. 이런 신경망의 초기 가중치에 따른 단점을 보안하기 위해 통계적 알고리즘의 접목을 통해 Hybrid된 신경망 보완 알고리즘을 제시하고자 하였다. 논문을 위한 기본 가정으로 신경망의 가장 기본인 SLP 알고리즘을 바탕으로 활성함수에 가장 일반적으로 사용되는 Sigmoid 활성함수를 이용하였을 때, 초기 가중치로 기존의 임의 난수 생성 방식이 아닌 통계적 로지스틱 회귀분석의 계수값(mle)을 제시하여 이를 초기치로 사용한 경우와 그렇지 않은 경우의 예측 정확성과 수렴의 Performance정도를 비교하여 가장 효과적인 초기치 방법을 제시하고자 하였다.

• Journal of Korean Society of Transportation
• /
• v.13 no.1
• /
• pp.113-123
• /
• 1995

버스의 크기에 따른 운행시격과 승객 대기시간의 Trade-off 는 대중교통 운영정책의 중요한 기본개념의 하나로서 여태까지 선진 구미의 여러 교통공학작들에 많은 연구가 수행되어져 왔다. 본 논문에서는 시스템 전체의 최적화(System Optimization )라는 측면에서 차량비를 포함한 버스회사의 총 운행경비와 버스 승객들의 대기시간과 주행시간을 금전으로 호나산한 사용자 경비를 합산한 총비용을 최소화 사키는 적정 버스크기(=좌석수)를 산정할 수 있는 모델식을 개발코자 하였다. 이를 위해 우선 수집된 자료들의 회귀분석을 통해 버스 운행경비와 버스크기와의 관계를 규명하였으며 이를 토대로 하여 버스 좌석수를 결정변수(Optimizable Variable)로 하는 총비용에 관한 목적함수식을 도출하였다. 또한 개발된 모형의 적절성을 검증하기 위해 미국 수도 워싱턴 지역에서의 교통자료를 인용하여 사례조사를 하였으며 이를 통해 본 연구에서 도출된 모형식의 실용성을 확인할 수 있었다. 추후 본 연구에서 개발된 수식들은 국내의 버스운행 여건과 실태를 잘 반영할 수 있도록 광범위한 자료조사를 통해 수정되어야 할 것이다.

• Annual Conference of KIPS
• /
• 2021.11a
• /
• pp.476-479
• /
• 2021

DBMS 파라미터 튜닝이란 데이터베이스에서 제공하는 다양한 파라미터의 값을 조율하여, 최적의 성능을 도출하는 과정이다. 데이터베이스 종류에 따라 파라미터 개수가 수십 개에서 수백 개로 다양하며, 각 기능이 모두 다르기 때문에 최적의 조합을 찾는 것은 쉽지 않다. 선행 연구에서는 BO 기법을 사용하여 적절한 파라미터 값을 추출했지만, 파라미터 개수에 비례하여 차원이 커지는 문제가 발생한다. 본 논문에서는 통계적으로 파라미터를 분류하여 탐색 공간을 줄인 다음 단계적으로 BO 를 수행하는 PBO 방식을 제안한다. 파라미터 값을 랜덤하게 할당하여 벤치마킹한 결과값을 군집화한 후, 각 군집별로 파라미터와의 연관성을 분석해 높은 상관관계를 가진 파라미터를 매칭시켜 분류한다. 제안하는 방법론을 검증하기 위하여 8 가지 회귀 모델과의 비교 실험을 통해 제안한 방법론의 우수성을 검증하였다.

• Advanced Industrial SCIence
• /
• v.3 no.2
• /
• pp.8-16
• /
• 2024

This paper shows the system of drug classification, the goal of this is to foretell the apt drug for the patients based on their demographic and physiological traits. The dataset consists of various attributes like Age, Sex, BP (Blood Pressure), Cholesterol Level, and Na_to_K (Sodium to Potassium ratio), with the objective to determine the kind of drug being given. The models used in this paper are K-Nearest Neighbors (KNN), Logistic Regression and Random Forest. Further to fine-tune hyper parameters using 5-fold cross-validation, GridSearchCV was used and each model was trained and tested on the dataset. To assess the performance of each model both with and without hyper parameter tuning evaluation metrics like accuracy, confusion matrices, and classification reports were used and the accuracy of the models without GridSearchCV was 0.7, 0.875, 0.975 and with GridSearchCV was 0.75, 1.0, 0.975. According to GridSearchCV Logistic Regression is the most suitable model for drug classification among the three-model used followed by the K-Nearest Neighbors. Also, Na_to_K is an essential feature in predicting the outcome.

• Journal of the Korea Convergence Society
• /
• v.11 no.5
• /
• pp.35-41
• /
• 2020

This study was conducted to develop the efficient system for starch hydrolysis suppression using rutin, quercetin and dietary fiber through the statistical mixture design. The three components were replaced with wheat flour at the level of 10% and the mixed gel with three components was characterized by in vitro starch digestion. The mixture design was applied by simplex-centroid experimental model. The quadratic model (R²=0.86) was well fitted and the obtained regression equation indicated that the significant positive effects was observed in the quercetin and fiber mixture. Based on the statistical results, the best mixing ratio of quercetin and fiber was 72: 28 that led to the lowest predicted glycemic index (pGI). Their interactions on the pGI of starch digestibility were clearly visualized in the 3D surface plot. These results suggested that the mixture of quercetin and fiber interact strongly with wheat flour, consequently retarding starch hydrolysis by 15%.

• Resources Recycling
• /
• v.21 no.3
• /
• pp.39-47
• /
• 2012

This study was conducted to optimize a mixing design of lightweight aerated concrete with the blast furnace slag(BFS) using Box-Behnken method, one of response surface designs. The lightweight aerated concrete with the BFS was made on the conditions of steam curing method at atmospheric pressure. The experimental factors were unit Water(W)/total powder($P_d$) ratio, BFS replacement percentage and Al powder addition based on the total powder (${P_d}^*$%). From the results of the response surface analysis, regression models for dried specific gravity and compressive strength of the lightweight aerated concrete were derived. When the target values for dried specific gravity and compressive strength of the lightweight aerated concrete were set at 0.72 and 4.42 MPa respectively, its optimized mixing conditions driven from the regression models were 0.62 of $W/P_d$ ratio, 35.5% of BFS replacement and 0.05% of Al powder addition. This experimental design model was found to be credible by measuring the dried specific gravity and compressive strength of the sample made from the above mixing conditions.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.10 no.6
• /
• pp.81-87
• /
• 2010

In order to calculate the accurate runoff from a basin, nonlinearity in the relationship between rainfall and runoff has to be considered. Many runoff calculation models assume the linearity in the relationship or are too complicated to be analyzed. Therefore, the storage function method has been used in the prediction of flood because of the simplicity of the model. The storage function method has five parameters with related to the basin and rainfall characteristics which can be estimated by the empirical trial and error method. To optimize these parameters, regression method or optimization techniques such as genetic algorithm have been used, however, it is not easy to optimize them because of the complexity of the method. In this study, the metamodel is proposed to estimate those model parameters. The metamodel is the combination of artificial neural network and genetic algorithm. The model is consisted of two stages. In the first stage, an artificial neural network is constructed using the given rainfall-runoff relationship. In the second stage, the parameters of the storage function method are estimated using genetic algorithm and the trained artificial neural network. The proposed metamodel is applied in the Peong Chang River basin and the results are presented.

• Journal of the Korean GEO-environmental Society
• /
• v.14 no.5
• /
• pp.57-63
• /
• 2013

In this study, the characteristics of degradation and mineralization of diazinon using a statistical approach based on Box-Behnken design (BBD, one of response surface method) was investigated in an E-beam process, and also the main factors with diazinon concentration ($X_1$), irradiatin intensity ($X_2$) and pH ($X_3$) which consisted of 3 levels in each factor was set up to determine the effects of factors and optimization. At first, effects of pH and diazinon concentration were investigated to determine the proper range of application on response surface method(RSM). In statistical approach, the regression analysis and analysis of variance (ANOVA) were applied to evaluate the quantitative comparison of each factors in order to obtain the effects were irradiation intensity>diazinon concentration>pH. The regression model predicted the optimization point using the response optimizer to consider the effects of operation conditions were $Y_1=81.73-5.58X_1+23.69X_2-14.23X{_2}^2+4.22X{_3}^2(R^2=99.7%)$, $Y_2=35.23-3.01X_1+10.79X_2-7.58X_2{^2}(R^2=97.9%)$ and 95.7% of diazinon degradation, 41.8% of TOC reduction at 12.75mg/L and 4.26kGy, respectively. The pH condition was not significantly affects on E-beam process than other advanced oxidation processes (AOPs).