• International Journal of Internet, Broadcasting and Communication
• /
• v.11 no.4
• /
• pp.37-42
• /
• 2019

In this paper, we explore the details of three classic data augmentation methods and two generative model based oversampling methods. The three classic data augmentation methods are random sampling (RANDOM), Synthetic Minority Over-sampling Technique (SMOTE), and Adaptive Synthetic Sampling (ADASYN). The two generative model based oversampling methods are Conditional Generative Adversarial Network (CGAN) and Wasserstein Generative Adversarial Network (WGAN). In imbalanced data, the whole instances are divided into majority class and minority class, where majority class occupies most of the instances in the training set and minority class only includes a few instances. Generative models have their own advantages when they are used to generate more plausible samples referring to the distribution of the minority class. We also adopt CGAN to compare the data augmentation performance with other methods. The experimental results show that WGAN-based oversampling technique is more stable than other approaches (RANDOM, SMOTE, ADASYN and CGAN) even with the very limited training datasets. However, when the imbalanced ratio is too small, generative model based approaches cannot achieve satisfying performance than the conventional data augmentation techniques. These results suggest us one of future research directions.

• Journal of Digital Convergence
• /
• v.17 no.9
• /
• pp.71-78
• /
• 2019

This study aims to identify good borrowers within the context of P2P lending. P2P lending is a growing platform that allows individuals to lend and borrow money from each other. Inherent in any loans is credit risk of borrowers and needs to be considered before any lending. Specifically in the context of P2P lending, traditional models fall short and thus this study aimed to rectify this as well as explore the problem of class imbalances seen within credit risk data sets. This study implemented an over-sampling technique known as Synthetic Minority Over-sampling Technique (SMOTE). To test our approach, we implemented five benchmarking classifiers such as support vector machines, logistic regression, k-nearest neighbor, random forest, and deep neural network. The data sample used was retrieved from the publicly available LendingClub dataset. The proposed SMOTE revealed significantly improved results in comparison with the benchmarking classifiers. These results should help actors engaged within P2P lending to make better informed decisions when selecting potential borrowers eliminating the higher risks present in P2P lending.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.9
• /
• pp.2961-2975
• /
• 2022

The Location-Based Service (LBS) is one of the most well-known services on the Internet. Positioning is the primary association with LBS services. This study proposes an intelligent LoRa-based positioning system, called AI@LBS, to provide accurate location data. The fingerprint mechanism with the clustering algorithm in unsupervised learning filters out signal noise and improves computing stability and accuracy. In this study, data noise is filtered using the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm, increasing the positioning accuracy from 95.37% to 97.38%. The problem of data imbalance is addressed using the SMOTE (Synthetic Minority Over-sampling Technique) technique, increasing the positioning accuracy from 97.38% to 99.17%. A field test in the NTUST campus (www.ntust.edu.tw) revealed that AI@LBS system can reduce average distance error to 0.48m.

• The Korean Journal of Applied Statistics
• /
• v.37 no.4
• /
• pp.395-410
• /
• 2024

Text data is usually made up of a wide variety of unique words. Even in standard text data, it is common to find tens of thousands of different words. In text data analysis, usually, each unique word is treated as a variable. Thus, text data can be regarded as a dataset with a large number of variables. On the other hand, in text data classification, we often encounter class label imbalance problems. In the cases of substantial imbalances, the performance of conventional classification models can be severely degraded. To improve the classification performance of support vector machines (SVM) for imbalanced data, algorithms such as the Synthetic Minority Over-sampling Technique (SMOTE) can be used. The SMOTE algorithm synthetically generates new observations for the minority class based on the k-Nearest Neighbors (kNN) algorithm. However, in datasets with a large number of variables, such as text data, errors may accumulate. This can potentially impact the performance of the kNN algorithm. In this study, we propose a method for enhancing prediction performance for the minority class of imbalanced text data. Our approach involves employing variable selection to generate new synthetic observations in a reduced space, thereby improving the overall classification performance of SVM.

• The Korean Journal of Applied Statistics
• /
• v.37 no.4
• /
• pp.455 -465
• /
• 2024

SMOTE (synthetic minority over-sampling technique) has been used the most as a solution to the problem of imbalanced data. SMOTE selects the nearest neighbor based on Euclidean distance. However, Euclidean distance has the disadvantage of not considering the correlation between variables. In particular, the Mahalanobis distance has the advantage of considering the covariance of variables. But if there are outliers, they usually influence calculating the Mahalanobis distance. To solve this problem, we use the Mahalanobis distance by estimating the covariance matrix using MCD (minimum covariance determinant). Then apply Mahalanobis distance based on MCD to SMOTE to create new data. Therefore, we showed that in most cases this method provided high performance indicators for classifying imbalanced data.

• Journal of the Korean Geotechnical Society
• /
• v.39 no.6
• /
• pp.5-12
• /
• 2023

Proactive assessment of landslide susceptibility is necessary for minimizing casualties. This study proposes a methodology for classifying the landslide safety factor using a classification algorithm based on machine learning techniques. The high-risk area model is adopted to perform the classification and eight geotechnical parameters are adopted as inputs. Four classification algorithms-namely decision tree, k-nearest neighbor, logistic regression, and random forest-are employed for comparing classification accuracy for the safety factors ranging between 1.2 and 2.0. Notably, a high accuracy is demonstrated in the safety factor range of 1.2~1.7, but a relatively low accuracy is obtained in the range of 1.8~2.0. To overcome this issue, the synthetic minority over-sampling technique (SMOTE) is adopted to generate additional data. The application of SMOTE improves the average accuracy by ~250% in the safety factor range of 1.8~2.0. The results demonstrate that SMOTE algorithm improves the accuracy of classification algorithms when applied to geotechnical data.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.517-517
• /
• 2023

현재 우리나라에서는 4대강 및 주요 호소 29지점을 대상으로 조류경보제가 시행되고 있으며 조류 경보 단계는 실시간 모니터링지점에서 측정되는 유해 조류의 셀농도를 기반으로 발령 단계가 결정된다. 상수원 구간은 관심, 경계, 조류 대발생, 해제 또는 미발생 총 4구간으로 구성되며, 친수 활동 구간의 경우 조류 대발생을 제외한 3구간으로 구성된다. 현재 시행되는 조류 경보제의 목적은 유해 조류 발생 시 사후 대응 방안 마련에 보다 초점이 맞춰져 있으며 특히, 모니터링 주기 확대 여부, 오염원 관리 방안 마련, 조류 제거 여부 등의 의사 결정 수단으로 사용되고 있다. 하지만 조류 경보 단계에 대한 사전 예측이 가능한 경우 유해 조류의 성장을 억제할 수 있으며 이를 통해 안전하고 깨끗한 수자원을 확보할 수 있다. 본 연구에서는 조류 경보 단계의 사전적 예측을 위해 국가 실시간 측정망에서 제공하는 전국 보 모니터링 종합 정보 자료, 기상측정망 자료, 실시간 보 현황 자료를 활용하여 예측 모델을 구축하였다. 또한, 단계 예측의 정확도를 개선하기 위해 변수 선택 기법을 활용하여 조류 경보 단계에 영향을 미치는 환경변수를 선정하였으며 자료의 불균형으로 인해 모델 학습 과정에서 발생하는 예측 오류를 최소화하기 위해 다양한 샘플링 기법을 적용하여 모델의 성능을 평가하였다. 변수 선택 및 샘플링 기법을 고려하지 않은 원자료를 사용하여 예측 모델을 구축한 결과 관심 단계(Level-1) 및 경보 단계(Level-2)에 대해 각각 50%, 62.5%의 예측 정확도를 보인 반면 비선형 변수 선택 기법 및 Synthetic Minority Over-sampling Technique-Edited Nearrest Neighbor(SMOTE-ENN) 샘플링 기법을 적용하여 구축한 모델에서는 Level-1은 85.7%, Level-2는 75.0%의 예측 정확도를 보였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.294-297
• /
• 2024

Malware detection has become increasingly critical with the proliferation of end devices. To improve detection rates and efficiency, the research focus in malware detection has shifted towards leveraging machine learning and deep learning approaches. This shift is particularly relevant in the context of the widespread adoption of end devices, including smartphones, Internet of Things devices, and personal computers. Machine learning techniques are employed to train models on extensive datasets and evaluate various features, while deep learning algorithms have been extensively utilized to achieve these objectives. In this research, we introduce TabNet, a novel architecture designed for deep learning with tabular data, specifically tailored for enhancing malware detection techniques. Furthermore, the Synthetic Minority Over-Sampling Technique is utilized in this work to counteract the challenges posed by imbalanced datasets in machine learning. SMOTE efficiently balances class distributions, thereby improving model performance and classification accuracy. Our study demonstrates that SMOTE can effectively neutralize class imbalance bias, resulting in more dependable and precise machine learning models.

• 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.

• Geomechanics and Engineering
• /
• v.37 no.6
• /
• pp.539-554
• /
• 2024

This study explores development of prediction model for seismic site classification through the integration of machine learning techniques with horizontal-to-vertical spectral ratio (HVSR) methodologies. To improve model accuracy, the research employs outlier detection methods and, synthetic minority over-sampling technique (SMOTE) for data balance, and evaluates using seven machine learning models using seismic data from KiK-net. Notably, light gradient boosting method (LGBM), gradient boosting, and decision tree models exhibit improved performance when coupled with SMOTE, while Multiple linear regression (MLR) and Support vector machine (SVM) models show reduced efficacy. Outlier detection techniques significantly enhance accuracy, particularly for LGBM, gradient boosting, and voting boosting. The ensemble of LGBM with the isolation forest and SMOTE achieves the highest accuracy of 0.91, with LGBM and local outlier factor yielding the highest F1-score of 0.79. Consistently outperforming other models, LGBM proves most efficient for seismic site classification when supported by appropriate preprocessing procedures. These findings show the significance of outlier detection and data balancing for precise seismic soil classification prediction, offering insights and highlighting the potential of machine learning in optimizing site classification accuracy.