• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.20 no.6
• /
• pp.1063-1068
• /
• 2016

Anomalous propagation echo is a kind of abnormal radar signal occurred by irregularly refracted radar beam caused by temperature or humidity. The echo frequently appears in ground-based weather radar due to its observation principle and disturb weather forecasting process. In order to improve accuracy of weather forecasting, it is important to analyze radar data precisely. Therefore, there are several ongoing researches about identifying the anomalous propagation echo with data mining techniques. This paper conducts researches about implementation of classification method which can separate the anomalous propagation echo in the raw radar data using naive Bayes classifier with various kinds of observation results. Considering that collected data has a class imbalanced problem, this paper includes SMOTE method. It is confirmed that the fine classification results are derived by the suggested classifier with balanced dataset using actual appearance cases of the echo.

• Journal of the Korea Convergence Society
• /
• v.12 no.1
• /
• pp.49-55
• /
• 2021

Recently, with the development of database, it is possible to store a lot of data generated in finance, security, and networks. These data are being analyzed through classifiers based on machine learning. The main problem at this time is data imbalance. When we train imbalanced data, it may happen that classification accuracy is degraded due to over-fitting with majority class data. To overcome the problem of data imbalance, oversampling strategy that increases the quantity of data of minority class data is widely used. It requires to tuning process about suitable method and parameters for data distribution. To improve the process, In this study, we propose a strategy to explore and optimize oversampling combinations and ratio based on various methods such as synthetic minority oversampling technique and generative adversarial networks through genetic algorithms. After sampling credit card fraud detection which is a representative case of data imbalance, with the proposed strategy and single oversampling strategies, we compare the performance of trained classifiers with each data. As a result, a strategy that is optimized by exploring for ratio of each method with genetic algorithms was superior to previous strategies.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.483-486
• /
• 2023

최근 금융 분야는 예측 모델의 복잡성으로 인한 블랙박스 문제와 금융 규제에 대한 관심이 높아지고 있다. 이에 따라 금융 업계는 신뢰성과 투명성을 강조하며, 특히 신용평가 분야에서 설명 가능한 모델 연구가 활발히 진행되고 있다. 또한, 해당 분야에서 소수 클래스에 대해 충분히 학습하지 못하고 다수 클래스에 과적합 될 수 있는 데이터 불균형 문제 역시 강조되고 있다. 이는 제 2종 오류(Type 2 Error)를 최소화해야 하는 상황에서 더욱 부각되며, 대출 상환 능력이 낮은 고객을 최대한 식별해야 하는 개인 신용평가 문제에서 매우 중요한 화두로 떠오르고 있다. 본 논문에서는 어텐션 메커니즘을 활용하여 모델의 설명 가능성을 개선하고, 분석 결과를 해석하는 데 도움이 되고자 한다. 더 나아가, SMOTE, GAN, ADASYN 등 총 다섯 가지 데이터 증강 기법을 실험하여, 이를 앙상블 하였을 때 소수 클래스 레이블에 대한 분류 정확도를 크게 개선할 수 있음을 확인하였다.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.12
• /
• pp.445-452
• /
• 2022

Class distribution of unbalanced data is an important part of the digital world and is a significant part of cybersecurity. Abnormal activity of unbalanced data should be found and problems solved. Although a system capable of tracking patterns in all transactions is needed, machine learning with disproportionate data, which typically has abnormal patterns, can ignore and degrade performance for minority layers, and predictive models can be inaccurately biased. In this paper, we predict target variables and improve accuracy by combining estimates using Synthetic Minority Oversampling Technique (SMOTE) and Light GBM algorithms as an approach to address unbalanced datasets. Experimental results were compared with logistic regression, decision tree, KNN, Random Forest, and XGBoost algorithms. The performance was similar in accuracy and reproduction rate, but in precision, two algorithms performed at Random Forest 80.76% and Light GBM 97.16%, and in F1-score, Random Forest 84.67% and Light GBM 91.96%. As a result of this experiment, it was confirmed that Light GBM's performance was similar without deviation or improved by up to 16% compared to five algorithms.

• Journal of Practical Engineering Education
• /
• v.16 no.1_spc
• /
• pp.59-64
• /
• 2024

As the aging population grows, the incidence of cancer is increasing. Skin cancer appears externally, but people often don't notice it or simply overlook it. As a result, if the early detection period is missed, the survival rate in the case of late stage cancer is only 7.5-11%. However, the disadvantage of diagnosing, serious skin cancer is that it requires a lot of time and money, such as a detailed examination and cell tests, rather than simple visual diagnosis. To overcome these challenges, we propose an Attention-based CNN model skin cancer classification system. If skin cancer can be detected early, it can be treated quickly, and the proposed system can greatly help the work of a specialist. To mitigate the problem of image data imbalance according to skin cancer type, this skin cancer classification model applies the Over Sampling, technique to data with a high distribution ratio, and adds a pre-learning model without an Attention layer. This model is then compared to the model without the Attention layer. We also plan to solve the data imbalance problem by strengthening data augmentation techniques for specific classes.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.6 s.44
• /
• pp.99-106
• /
• 2006

Type classification is a very needed step in recognizing huge character set language such as korean characters. Since most previous researches are based on the composition rule of Korean characters, it has been difficult to correctly classify composite vowel characters and problem space was not divided equally for the lack of classification of last consonant which is relatively bigger than other graphemes. In this paper, I Propose a new type classification method in which horizontal vowel is extracted before vortical vowel and last consonants are further classified into one of five small groups based on horizontal projection profile. The new method uses 19 character types which is more stable than previous 6 types or 15 types. Through experiments on 1.000 frequently used character sets and 30.614 characters scanned from several magazines, I showed that the proposed method is more useful classifying Korean characters of huge set.

• Information Systems Review
• /
• v.24 no.4
• /
• pp.97-118
• /
• 2022

As generative adversarial network (GAN) based oversampling techniques have achieved impressive results in class imbalance of unstructured dataset such as image, many studies have begun to apply it to solving the problem of imbalance in structured dataset. However, these studies have failed to reflect the characteristics of structured data due to changing the data structure into an unstructured data format. In order to overcome the limitation, this study adapted CycleGAN to reflect the characteristics of structured data, and proposed hybridization of synthetic minority oversampling technique (SMOTE) and the adapted CycleGAN. In particular, this study tried to overcome the limitations of existing studies by using a one-dimensional convolutional neural network unlike previous studies that used two-dimensional convolutional neural network. Oversampling based on the method proposed have been experimented using various datasets and compared the performance of the method with existing oversampling methods such as SMOTE and adaptive synthetic sampling (ADASYN). The results indicated the proposed hybrid oversampling method showed superior performance compared to the existing methods when data have more dimensions or higher degree of imbalance. This study implied that the classification performance of oversampling structured data can be improved using the proposed hybrid oversampling method that considers the characteristic of structured data.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.687-689
• /
• 2024

위암은 전 세계적인 주요 건강문제이며, 근치적 위절제술은 위암의 표준치료이다. 근치적 위절제술 후 치료목표는 합병증 발생을 낮춰 병전 상태로 빠르게 회복하는 데 있다. 따라서, 근치적 위절제술 후 합병증 발생 여부를 선별하여 예측할 수 있는 성능이 좋은 모델을 개발하는 것은 위암환자의 회복에 매우 중요하다. 랜덤포레스트 모델은 여러 개의 결정트리를 활용한 배깅 방식의 대표적인 알고리즘으로 의료 데이터를 기반으로 한 예측에 있어 뛰어난 성능을 보여 주었다. 그러나 실제 데이터는 불균형이 빈번하게 발생하여 모델의 예측 성능에 영향을 미치므로, 최적의 분류 임계값을 설정하여 다수 클래스에 대한 편향을 줄이는 것이 중요하다. 따라서, 본 연구는 최근 10년 간 일개 대학병원의 전자의무기록 데이터를 활용하여 근치적 위절제술 후 합병증 발생을 예측하는 랜덤포레스트 모델을 개발하고, 임계값 설정을 통해 불균형 데이터에 대한 모델의 성능을 평가하고자 한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.05a
• /
• pp.401-404
• /
• 2022

한국 문학이 세계로 뻗어나감에 따라 해외 시장에서 자리를 잡는 것이 중요해진 시점이다. 본 연구에서는 2016 년도부터 2020 년도까지 최근 5 년간 해외 출간된 도서들 중에서 굿셀러로 분류되는 누적 5 천부 이상 판매 여부를 예측하고자 했다. 굿셀러로 분류되는 도서는 전체 번역 도서 중 적은 비율을 차지하여 데이터 불균형이 발생하였으며, 본 연구에서는 SMOTE 기법과 앙상블 알고리즘을 적용하여 데이터 불균형 문제를 해결하였다. 그 결과, 데이터 클래스 비율이 1:1 에 가까울수록 성능 개선 효과가 나타났으며 LightGBM 모델이 99.83%의 AUC 값을 얻어 다른 앙상블 알고리즘에 비해 가장 좋은 예측 성능을 보임을 검증하였다. 또한 누적 5 천부 이상 판매 여부 예측에 있어 큰 영향을 미치는 변수로는 작가가 가장 중요한 요인으로 나타났으며 출간 국가, 그리고 평점 평균, 평점 참여자 수 같은 온라인 요인도 판매 예측에 유의미한 변수로 나타난 것을 확인할 수 있었다.

• The KIPS Transactions:PartD
• /
• v.19D no.1
• /
• pp.21-28
• /
• 2012

Predication of protein interaction sites for monomer structures can reduce the search space for protein docking and has been regarded as very significant for predicting unknown functions of proteins from their interacting proteins whose functions are known. In the other hand, the prediction of interaction sites has been limited in crystallizing weakly interacting complexes which are transient and do not form the complexes stable enough for obtaining experimental structures by crystallization or even NMR for the most important protein-protein interactions. This work reports the calculation of 3D surface patches of complex structures and their properties and a machine learning approach to build a predictive model for the 3D surface patches in interaction and non-interaction sites using support vector machine. To overcome classification problems for class imbalanced data, we employed an under-sampling technique. 9 properties of the patches were calculated from amino acid compositions and secondary structure elements. With 10 fold cross validation, the predictive model built from SVM achieved an accuracy of 92.7% for classification of 3D patches in interaction and non-interaction sites from 147 complexes.