• Ecology and Resilient Infrastructure
• /
• v.10 no.4
• /
• pp.107-115
• /
• 2023

High turbidity in source water can have adverse effects on water treatment plant operations and aquatic ecosystems, necessitating turbidity management. Consequently, research aimed at predicting river turbidity continues. This study developed a multi-class classification model for prediction of turbidity using LightGBM (Light Gradient Boosting Machine), a representative ensemble machine learning algorithm. The model utilized data that was classified into four classes ranging from 1 to 4 based on turbidity, from low to high. The number of input data points used for analysis varied among classes, with 945, 763, 95, and 25 data points for classes 1 to 4, respectively. The developed model exhibited precisions of 0.85, 0.71, 0.26, and 0.30, as well as recalls of 0.82, 0.76, 0.19, and 0.60 for classes 1 to 4, respectively. The model tended to perform less effectively in the minority classes due to the limited data available for these classes. To address data imbalance, the SMOTE (Synthetic Minority Over-sampling Technique) algorithm was applied, resulting in improved model performance. For classes 1 to 4, the Precision and Recall of the improved model were 0.88, 0.71, 0.26, 0.25 and 0.79, 0.76, 0.38, 0.60, respectively. This demonstrated that alleviating data imbalance led to a significant enhancement in Recall of the model. Furthermore, to analyze the impact of differences in input data composition addressing the input data imbalance, input data was constructed with various ratios for each class, and the model performances were compared. The results indicate that an appropriate composition ratio for model input data improves the performance of the machine learning model.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.20
• /
• pp.8719-8724
• /
• 2014

Background: Breast cancer incidence is increasing among South Asian migrants to the United States (US). However, their utilization of cancer screening services is poor. This study characterizes attitudes of South Asians towards breast health and screening in a community sample. Materials and Methods: A cross-sectional survey based on the Health Belief Model (HBM) was conducted among South Asians (n=124) in New Jersey and Chicago. The following beliefs and attitudes towards breast cancer screening were assessed-health motivation, breast self-examination confidence, breast cancer susceptibility and fear, and mammogram benefits and barriers. Descriptive statistics and Spearman rank correlation coefficients were computed for HBM subscales. Findings: Mean age of participants was 36 years with an average 10 years stay in the US. Most women strived to care for their health ($3.82{\pm}1.18$) and perceived high benefits of screening mammography ($3.94{\pm}0.95$). However, they perceived lower susceptibility to breast cancer in the future ($2.30{\pm}0.94$). Conclusions: Increasing awareness of breast cancer risk for South Asian women may have a beneficial effect on cancer incidence because of their positive attitudes towards health and breast cancer screening. This is especially relevant because South Asians now constitute one of the largest minority populations in the US and their incidence of breast cancer is steadily increasing.

• Proceedings of the KIPE Conference
• /
• 2001.10a
• /
• pp.26-30
• /
• 2001

In this paper, transient characteristics of IGBT has been analytically solved to express the excess minority carrier distribution in active base region and the output voltage. Non-Punch Through(NPT) structure has been selected to prove the validity of the model. It is based on the equivalent circuit of MOSFET which supplies a low gain and a high level injection to the base of BJT. None of the quasi static conditions have been assumed to trace the transient characteristics. The basic elements of the model have been derived from the ambipolar transport theory. Theoretical predictions of the output voltages have been obtained with different lifetimes and compared with experimental and theoretical results available in the literature. From the analytical approach, good agreement has been obtained to provide reliable and fast output of the device.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.34 no.4
• /
• pp.73-89
• /
• 2009

Circular shareholdings among three companies are formed when company A owns stock in company B, company B owns stock in company C, and company C owns stock in company A. Since circular shareholdings among large family-controlled firms are used to give the controlling shareholder greater control or more opportunities to expropriate minority investors, the government has encouraged large business groups to gradually remove their circular shareholdings. In this paper, we propose a combinatorial optimization model that can answer the question, which equity investments among complicated investment relationships of one large business group should be removed to resolve its circular shareholdings. To the best knowledge of the authors, our research is the first one that has approached the circular shareholding problem in respect of management science. The proposed combinatorial optimization model are formulated into integer programming problem and applied to some Korean major business groups.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.13 no.2
• /
• pp.120-124
• /
• 2000

In this paper, the effects of trap distribution on switching characteristis of a lateral insulated gate bipolar transistor (LIGBT) are investigated. The simulations are performed in order to to analyze the effect of the positon, width and concentration of trap distribution model with a reduced minority carrier lifetime using 2D device simulator MEDICI. The turn off time for the proposed LIGBT model A with the trap injection is 0.8$mutextrm{s}$. These results indicate the improvement of about 2 times compared with the conventional LIGBT. It is shown that the trap distribution model is very effective to reduce the turn-off time with a little increasing of on-state voltage drop.

• International journal of advanced smart convergence
• /
• v.12 no.1
• /
• pp.53-58
• /
• 2023

With the fast development of artificial intelligence day by day, users are demanding explanations about the results of algorithms and want to know what parameters influence the results. In this paper, we propose a model for bankruptcy prediction with interpretability using the SHAP framework. SHAP (SHAPley Additive exPlanations) is framework that gives a visualized result that can be used for explanation and interpretation of machine learning models. As a result, we can describe which features are important for the result of our deep learning model. SHAP framework Force plot result gives us top features which are mainly reflecting overall model score. Even though Fully Connected Neural Networks are a "black box" model, Shapley values help us to alleviate the "black box" problem. FCNNs perform well with complex dataset with more than 60 financial ratios. Combined with SHAP framework, we create an effective model with understandable interpretation. Bankruptcy is a rare event, then we avoid imbalanced dataset problem with the help of SMOTE. SMOTE is one of the oversampling technique that resulting synthetic samples are generated for the minority class. It uses K-nearest neighbors algorithm for line connecting method in order to producing examples. We expect our model results assist financial analysts who are interested in forecasting bankruptcy prediction of companies in detail.

• International Journal of Advanced Culture Technology
• /
• v.10 no.4
• /
• pp.499-510
• /
• 2022

This study examines the reporting factors of crime against business in Korea and proposes a corresponding predictive model using machine learning. While many previous studies focused on the individual factors of theft victims, there is a lack of evidence on the reporting factors of crime against a business that serves the public good as opposed to those that protect private property. Therefore, we proposed a crime prevention model for the willingness factor of theft reporting in businesses. This study used data collected through the 2015 Commercial Crime Damage Survey conducted by the Korea Institute for Criminal Policy. It analyzed data from 834 businesses that had experienced theft during a 2016 crime investigation. The data showed a problem with unbalanced classes. To solve this problem, we jointly applied the Synthetic Minority Over Sampling Technique and the Tomek link techniques to the training data. Two prediction models were implemented. One was a statistical model using logistic regression and elastic net. The other involved a support vector machine model, tree-based machine learning models (e.g., random forest, extreme gradient boosting), and a stacking model. As a result, the features of theft price, invasion, and remedy, which are known to have significant effects on reporting theft offences, can be predicted as determinants of such offences in companies. Finally, we verified and compared the proposed predictive models using several popular metrics. Based on our evaluation of the importance of the features used in each model, we suggest a more accurate criterion for predicting var.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2004.07a
• /
• pp.405-408
• /
• 2004

In this work, Analytical model for voltage and current characteristics of NPT(Non-PunchThrough) IGBT(Insulated Gate Bipolar Transistor) was represented. voltage and current characteristics models were based on prediction on power loss of NPT IGBT during transient condition. For Analytical current model, excess carrier concentration and accumulated charge in active base width was analyzed with time variance. Analytical models were simulated by varying lifetime of excess minority carrier.

• Proceedings of the Korea Society for Simulation Conference
• /
• 2003.11a
• /
• pp.75-85
• /
• 2003

With recent advances in technology and the changing nature of business, call center management has become a rapidly growing industry. However theoretical analysis about the call center system is very difficult, and the forecasting of call volume also. In the situation, it is significant that we study call-flow system, design system model and perform simulation. If these are possible, it is able to control the staff schedule and the resource management efficiently. This study introduces the process of applying the call center to simulation. So, it is feasible to break from the intuitive management by a minority manager and analyze it scientifically. The enterprise can reduce unnecessary expense, make an offer high quality to user in a keen competition

• ETRI Journal
• /
• v.15 no.3
• /
• pp.61-69
• /
• 1994

A new analytical model for the base current of Si/SiGe/Si heterojunction bipolar transistors(HBTs) has been developed. This model includes the hole injection current from the base to the emitter, and the recombination components in the space charge region(SCR) and the neutral base. Distinctly different from other models, this model includes the following effects on each base current component by using the boundary condition of the excess minority carrier concentration at SCR boundaries: the first is the effect of the parasitic potential barrier which is formed at the Si/SiGe collector-base heterojunction due to the dopant outdiffusion from the SiGe base to the adjacent Si collector, and the second is the Ge composition grading effect. The effectiveness of this model is confirmed by comparing the calculated result with the measured plot of the base current vs. the collector-base bias voltage for the ungraded HBT. The decreasing base current with the increasing the collector-base reverse bias voltage is successfully explained by this model without assuming the short-lifetime region close to the SiGe/Si collector-base junction, where a complete absence of dislocations is confirmed by transmission electron microscopy (TEM)[1].The recombination component in the neutral base region is shown to dominate other components even for HBTs with a thin base, due to the increased carrier storage in the vicinity of the parasitic potential barrier at collector-base heterojunction.