A5093887444- Anomaly Detection Techniques and Applications
- Fault Detection and Control Systems
- Advanced Statistical Methods and Models
- Image Processing Techniques and Applications
- Imbalanced Data Classification Techniques
Karlsruhe Institute of Technology
2024
Outlier detection in high-dimensional tabular data is an important task mining, essential for many downstream tasks and applications. Existing unsupervised outlier algorithms face one or more problems, including inlier assumption (IA), curse of dimensionality (CD), multiple views (MV). To address these issues, we introduce Generative Subspace Adversarial Active Learning (GSAAL), a novel approach that uses Network with adversaries. These adversaries learn the marginal class probability...
Experimental studies are a cornerstone of machine learning (ML) research. A common, but often implicit, assumption is that the results study will generalize beyond itself, e.g. to new data. That is, there high probability repeating under different conditions yield similar results. Despite importance concept, problem measuring generalizability remains open. This probably due lack mathematical formalization experimental studies. In this paper, we propose such and develop quantifiable notion...
Outlier generation is a popular technique used to solve important outlier detection tasks. Generating outliers with realistic behavior challenging. Popular existing methods tend disregard the “multiple views” property of in high-dimensional spaces.The only method accounting for this falls short efficiency and effectiveness. We propose Bisect , new that creates mimicking said property. To do so, employs novel proposition introduced article stating how efficiently generate outliers. Our has...
Outlier generation is a popular technique used for solving important outlier detection tasks. Generating outliers with realistic behavior challenging. Popular existing methods tend to disregard the 'multiple views' property of in high-dimensional spaces. The only method accounting this falls short efficiency and effectiveness. We propose BISECT, new that creates mimicking said property. To do so, BISECT employs novel proposition introduced article stating how efficiently generate outliers....