• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2024년도 춘계학술발표대회
• /
• 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.

• 정보관리학회지
• /
• 제36권2호
• /
• pp.57-77
• /
• 2019

대표적인 앙상블 기법으로서 랜덤포레스트(RF)를 문헌정보학 분야의 학술지 논문에 대한 자동분류에 적용하였다. 특히, 국내 학술지 논문에 주제 범주를 자동 할당하는 분류 성능 측면에서 트리 수, 자질선정, 학습집합 크기 등 주요 요소들에 대한 다각적인 실험을 수행하였다. 이를 통해, 실제 환경의 불균형 데이터세트(imbalanced dataset)에 대하여 랜덤포레스트(RF)의 성능을 최적화할 수 있는 방안을 모색하였다. 결과적으로 국내 학술지 논문의 자동분류에서 랜덤포레스트(RF)는 트리 수 구간 100~1000(C)과 카이제곱통계량(CHI)으로 선정한 소규모의 자질집합(10%), 대부분의 학습집합(9~10년)을 사용하는 경우에 가장 좋은 분류 성능을 기대할 수 있는 것으로 나타났다.

• 스마트미디어저널
• /
• 제8권2호
• /
• pp.46-57
• /
• 2019

In recent research, deep learning-based methods have achieved state-of-the-art performance in various computer vision tasks. However, these methods are commonly supervised, and require huge amounts of annotated data to train. Acquisition of data demands an additional costly effort, particularly for the tasks where it becomes challenging to obtain large amounts of data considering the time constraints and the requirement of professional human diligence. In this paper, we present a data level synthetic sampling solution to learn from small and imbalanced data sets using Generative Adversarial Networks (GANs). The reason for using GANs are the challenges posed in various fields to manage with the small datasets and fluctuating amounts of samples per class. As a result, we present an approach that can improve learning with respect to data distributions, reducing the partiality introduced by class imbalance and hence shifting the classification decision boundary towards more accurate results. Our novel method is demonstrated on a small dataset of 2789 tomato plant disease images, highly corrupted with class imbalance in 9 disease categories. Moreover, we evaluate our results in terms of different metrics and compare the quality of these results for distinct classes.

• International Journal of Computer Science & Network Security
• /
• 제22권10호
• /
• pp.1-10
• /
• 2022

Nowadays software defect prediction (SDP) is most active research going on in software engineering. Early detection of defects lowers the cost of the software and also improves reliability. Machine learning techniques are widely used to create SDP models based on programming measures. The majority of defect prediction models in the literature have problems with class imbalance and high dimensionality. In this paper, we proposed Centroid and Nearest Neighbor based Class Imbalance Reduction (CNNCIR) technique that considers dataset distribution characteristics to generate symmetry between defective and non-defective records in imbalanced datasets. The proposed approach is compared with SMOTE (Synthetic Minority Oversampling Technique). The high-dimensionality problem is addressed using Ant Colony Optimization (ACO) technique by choosing relevant features. We used nine different classifiers to analyze six open-source software defect datasets from the PROMISE repository and seven performance measures are used to evaluate them. The results of the proposed CNNCIR method with ACO based feature selection reveals that it outperforms SMOTE in the majority of cases.

• 지능정보연구
• /
• 제26권1호
• /
• pp.23-45
• /
• 2020

기업의 경쟁력 확보를 위해 판별 알고리즘을 활용한 의사결정 역량제고가 필요하다. 하지만 대부분 특정 문제영역에는 적합한 판별 알고리즘이 어떤 것인지에 대한 지식은 많지 않아 대부분 시행착오 형식으로 최적 알고리즘을 탐색한다. 즉, 데이터셋의 특성에 따라 어떠한 분류알고리즘을 채택하는 것이 적합한지를 판단하는 것은 전문성과 노력이 소요되는 과업이었다. 이는 메타특징(Meta-Feature)으로 불리는 데이터셋의 특성과 판별 알고리즘 성능과의 연관성에 대한 연구가 아직 충분히 이루어지지 않았기 때문이며, 더구나 다중 클래스(Multi-Class)의 특성을 반영하는 메타특징에 대한 연구 또한 거의 이루어진 바 없다. 이에 본 연구의 목적은 다중 클래스 데이터셋의 메타특징이 판별 알고리즘의 성능에 유의한 영향을 미치는지에 대한 실증 분석을 하는 것이다. 이를 위해 본 연구에서는 다중 클래스 데이터셋의 메타특징을 데이터셋의 구조와 데이터셋의 복잡도라는 두 요인으로 분류하고, 그 안에서 총 7가지 대표 메타특징을 선택하였다. 또한, 본 연구에서는 기존 연구에서 사용하던 IR(Imbalanced Ratio) 대신 시장집중도 측정 지표인 허핀달-허쉬만 지수(Herfindahl-Hirschman Index, HHI)를 메타특징에 포함하였으며, 역ReLU 실루엣 점수(Reverse ReLU Silhouette Score)도 새롭게 제안하였다. UCI Machine Learning Repository에서 제공하는 복수의 벤치마크 데이터셋으로 다양한 변환 데이터셋을 생성한 후에 대표적인 여러 판별 알고리즘에 적용하여 성능 비교 및 가설 검증을 수행하였다. 그 결과 대부분의 메타특징과 판별 성능 사이의 유의한 관련성이 확인되었으며, 일부 예외적인 부분에 대한 고찰을 하였다. 본 연구의 실험 결과는 향후 메타특징에 따른 분류알고리즘 추천 시스템에 활용할 것이다.

• ETRI Journal
• /
• 제44권2호
• /
• pp.327-338
• /
• 2022

Classification models for human action recognition require robust features and large training sets for good generalization. However, data augmentation methods are employed for imbalanced training sets to achieve higher accuracy. These samples generated using data augmentation only reflect existing samples within the training set, their feature representations are less diverse and hence, contribute to less precise classification. This paper presents new data augmentation and action representation approaches to grow training sets. The proposed approach is based on two fundamental concepts: virtual video generation for augmentation and representation of the action videos through robust features. Virtual videos are generated from the motion history templates of action videos, which are convolved using a convolutional neural network, to generate deep features. Furthermore, by observing an objective function of the genetic algorithm, the spatiotemporal features of different samples are combined, to generate the representations of the virtual videos and then classified through an extreme learning machine classifier on MuHAVi-Uncut, iXMAS, and IAVID-1 datasets.

• International Journal of Internet, Broadcasting and Communication
• /
• 제15권3호
• /
• pp.66-72
• /
• 2023

Little attention appears to have been paid to the relevance of learning a good representation function in solving long tail tasks. Therefore, we propose a new loss function to ensure a good representation is learnt while learning to classify. We call this loss function Triplet Class-Wise Difficulty-Based (TriCDB-CE) Loss. It is a combination of the Triplet Loss and Class-wise Difficulty-Based Cross-Entropy (CDB-CE) Loss. We prove its effectiveness empirically by performing experiments on three benchmark datasets. We find improvement in accuracy after comparing with some baseline methods. For instance, in the CIFAR-10-LT, 7 percentage points (pp) increase relative to the CDB-CE Loss was recorded. There is more room for improvement on Places-LT.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 춘계학술발표대회
• /
• pp.585-587
• /
• 2023

Long-tailed datasets have an imbalanced distribution because they consist of a different number of data samples for each class. However, there are problems of the performance degradation in tail-classes and class-accuracy imbalance for all classes. To address these problems, this paper suggests a learning method for training of long-tailed dataset. The proposed method uses and combines two methods; one is a resampling method to generate a uniform mini-batch to prevent the performance degradation in tail-classes, and the other is a reweighting method to address the accuracy imbalance problem. The purpose of our proposed method is to train the learning models to have uniform accuracy for each class in a long-tailed dataset.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 추계학술발표대회
• /
• pp.605-607
• /
• 2023

Recent research has highlighted the effectiveness of Deep Learning (DL) techniques in automating the detection of lung sound anomalies. However, the available lung sound datasets often suffer from limitations in both size and balance, prompting DL methods to employ data preprocessing such as augmentation and transfer learning techniques. These strategies, while valuable, contribute to the increased complexity of DL models and necessitate substantial training memory. In this study, we proposed a streamlined and lightweight DL method but effectively detects lung sound anomalies from small and imbalanced dataset. The utilization of 1D dilated convolutional neural networks enhances sensitivity to lung sound anomalies by efficiently capturing deep temporal features and small variations. We conducted a comprehensive evaluation of the ICBHI dataset and achieved a notable improvement over state-of-the-art results, increasing the average score of sensitivity and specificity metrics by 2.7%.

• International Journal of Computer Science & Network Security
• /
• 제23권12호
• /
• pp.1-12
• /
• 2023

Ensuring the security of Supervisory Control and Data Acquisition (SCADA) and Industrial Control Systems (ICS) is paramount to safeguarding the reliability and safety of critical infrastructure. This paper addresses the significant threat posed by reconnaissance attacks on SCADA/ICS networks and presents an innovative methodology for enhancing their protection. The proposed approach strategically employs imbalance dataset handling techniques, ensemble methods, and feature engineering to enhance the resilience of SCADA/ICS systems. Experimentation and analysis demonstrate the compelling efficacy of our strategy, as evidenced by excellent model performance characterized by good precision, recall, and a commendably low false negative (FN). The practical utility of our approach is underscored through the evaluation of real-world SCADA/ICS datasets, showcasing superior performance compared to existing methods in a comparative analysis. Moreover, the integration of feature augmentation is revealed to significantly enhance detection capabilities. This research contributes to advancing the security posture of SCADA/ICS environments, addressing a critical imperative in the face of evolving cyber threats.