SECODA: Segmentation- and Combination-Based Detection of Anomalies
UWV（荷兰乌得勒支工作与收入机构）
Ralph Foorthuis
备注：
12 pages (including DSAA conference poster), 9 figures, 3 tables. Presented at DSAA 2017, the IEEE International Conference on Data Science and Advanced Analytics
This study introduces SECODA, a novel general-purpose unsupervised non-parametric anomaly detection algorithm for datasets containing continuous and categorical attributes. The method is guaranteed to identify cases with unique or sparse combinations of attribute values. Continuous attributes are discretized repeatedly in order to correctly determine the frequency of such value combinations. The concept of constellations, exponentially increasing weights and discretization cut points, as well as a pruning heuristic are used to detect anomalies with an optimal number of iterations. Moreover, the algorithm has a low memory imprint and its runtime performance scales linearly with the size of the dataset. An evaluation with simulated and real-life datasets shows that this algorithm is able to identify many different types of anomalies, including complex multidimensional instances. An evaluation in terms of a data quality use case with a real dataset demonstrates that SECODA can bring relevant and practical value to real-world settings.
Automated anomaly-aware 3D segmentation of bones and cartilages in knee MR images from the Osteoarthritis Initiative
The University of Queensland（昆士兰大学）；Australian e-Health Research Centre（澳大利亚电子健康研究中心）；Commonwealth Scientific and Industrial Research Organisation（联邦科学与工业研究组织）
Reversing the Abnormal: Pseudo-Healthy Generative Networks for Anomaly Detection
Technical University of Munich（慕尼黑工业大学）；Helmholtz AI（亥姆霍兹人工智能）；Helmholtz Center Munich（亥姆霍兹慕尼黑中心）；Klinikum Rechts der Isar（伊萨尔河右岸医院）；Imperial College London（帝国理工学院）；King’s College London（伦敦国王学院）
Exploiting Epistemic Uncertainty of Anatomy Segmentation for Anomaly Detection in Retinal OCT
Medical University Vienna（维也纳医科大学）；University of Vienna（维也纳大学）；NVIDIA Corporation（英伟达公司）
通过检测医学图像中的相关生物标志物来辅助诊断和治疗指导。虽然监督式深度学习可以对病理区域进行准确分割，但它受到限制，需要对这些区域进行先验定义，大规模注释以及训练集中的代表性患者队列。相反，异常检测不限于病理的特定定义，并且允许在没有注释的情况下对健康样本进行训练。然后，异常区域可以作为生物标志物发现的候选者。关于正常解剖结构的知识带来了用于检测异常的隐含信息。我们建议使用baidu深度学习来利用这一特性，基于这样的假设，即认知的不确定性将与正常训练集的解剖偏差相关。贝叶斯U形网是在一个定义良好的健康环境中训练的，使用现有方法产生的健康解剖结构的弱标签。在测试时，我们使用Monte Carlo dropout捕获模型的认知不确定性估计。一种新的后处理技术，然后应用到利用这些估计，并将其分层的外观平滑斑点形状的分割的异常。我们实验验证了这种方法在视网膜光学相干断层扫描（OCT）图像，使用弱标签的视网膜层。我们的方法在年龄相关性黄斑变性（AMD）病例的独立异常测试集中获得了0.789的Dice指数。所得到的分割允许非常高的准确性，以分离健康和患病的情况下，晚期湿性AMD，干地图状萎缩（GA），糖尿病性黄斑水肿（DME）和视网膜静脉阻塞（RVO）。最后，我们定性地观察到，我们的方法也可以检测正常扫描中的其他偏差，如切割边缘伪影。
IterMask3D: Unsupervised Anomaly Detection and Segmentation with Test-Time Iterative Mask Refinement in 3D Brain MR
Department of Engineering Science, University of Oxford, UK（工程科学系，牛津大学）；Department of Computer Science, Hong Kong Baptist University（计算机科学系，香港 Baptist 大学）；Nuffield Department of Clinical Neurosciences, University of Oxford（临床神经科学系，牛津大学）；Department of Neuroradiology, Oxford University Hospitals NHS Foundation Trust（放射学系，牛津大学医院 NHS 基础研究信托）
An Unsupervised Time Series Anomaly Detection Approach for Efficient Online Process Monitoring of Additive Manufacturing
College of Engineering and Computer Science, University of Texas Rio Grande Valley（工程与计算机科学学院，德克萨斯理工大学Rio Grande Valley分校）；School of Industrial Engineering and Management, Oklahoma State University（工业工程与管理学院，俄克拉荷马州立大学）
LLmFPCA-detect: LLM-powered Multivariate Functional PCA for Anomaly Detection in Sparse Longitudinal Texts
H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology（乔治亚理工学院H.Milton Stewart工业与系统工程学院）；AT&T Chief Data Office（AT&T首席数据办公室）；Department of Data Sciences and Operations, Marshall School of Business, University of Southern California（南加州大学马歇尔商学院数据科学与运营学院）
Machine Learning Group, TU Berlin（柏林技术大学机器学习组）；BIFOLD, Berlin, Germany（柏林BIFOLD）；Dept. of AI, Korea University, Seoul, Republic of Korea（韩国大学人工智能系）；MPI for Informatics, Saarbrücken, Germany（信息学Max Planck研究所）
