Robust Statistical Scaling of Outlier Scores: Improving the Quality of Outlier Probabilities for Outliers (Extended Version)
Johannes Gutenberg University Mainz(美因茨约翰内斯古滕贝格大学);University of Southern Denmark(南丹麦大学)
Philipp Röchner, Henrique O. Marques, Ricardo J. G. B. Campello, Arthur Zimek, Franz Rothlauf
备注:
15 pages, 4 figures, extended version of an original article published in Similarity Search and Applications. SISAP 2024. Lecture Notes in Computer Science, vol 15268. Springer, by Springer Nature
Outlier detection algorithms typically assign an outlier score to each observation in a dataset, indicating the degree to which an observation is an outlier. However, these scores are often not comparable across algorithms and can be difficult for humans to interpret. Statistical scaling addresses this problem by transforming outlier scores into outlier probabilities without using ground-truth labels, thereby improving interpretability and comparability across algorithms. However, the quality of this transformation can be different for outliers and inliers. Missing outliers in scenarios where they are of particular interest - such as healthcare, finance, or engineering - can be costly or dangerous. Thus, ensuring good probabilities for outliers is essential. This paper argues that statistical scaling, as commonly used in the literature, does not produce equally good probabilities for outliers as for inliers. Therefore, we propose robust statistical scaling, which uses robust estimators to improve the probabilities for outliers. We evaluate several variants of our method against other outlier score transformations for real-world datasets and outlier detection algorithms, where it can improve the probabilities for outliers.
Tabular Data Adapters: Improving Outlier Detection for Unlabeled Private Data
German Research Center for Artificial Intelligence (DFKI)(德国人工智能研究中心(DFKI));RPTU Kaiserslautern-Landau, Germany(鲁尔大学(RPTU)科布伦茨-兰道,德国);Hochschule Bonn-Rhein-Sieg (H-BRS)(波恩-莱茵-锡格应用技术大学(H-BRS))
ECOD: Unsupervised Outlier Detection Using Empirical Cumulative Distribution Functions
Northeastern University-Toronto(东北大学多伦多分校);Arima Inc.(Arima公司);Heinz College of Information Systems and Public Policy, Carnegie Mellon University(卡内基梅隆大学海因茨信息系统与公共政策学院);Potsdam Institute for Climate Impact Research(波茨坦气候影响研究所);Deggendorf Institute of Technology(代根多夫应用技术大学)
Fuzzy Granule Density-Based Outlier Detection with Multi-Scale Granular Balls
College of Computer Science and Software Engineering, Shenzhen University(深圳大学计算机与软件学院);National Engineering Laboratory for Big Data System Computing Technology, Shenzhen University(深圳大学大数据系统计算技术国家工程实验室);School of Artificial Intelligence and Computer Science, Nantong University(南通大学人工智能与计算机科学学院);Faculty of Data Science, City University of Macau(澳门城市大学数据科学学院);Department of Electrical and Computer Engineering, University of Alberta(阿尔伯塔大学电气与计算机工程系);Systems Research Institute, Polish Academy of Sciences(波兰科学院系统研究所);Department of Electrical and Computer Engineering, King Abdulaziz University(阿卜杜勒阿齐兹大学电气与计算机工程系);Department of Computer Engineering, Istinye University(伊斯提尼耶大学计算机工程系)
Augmenting The Weather: A Hybrid Counterfactual-SMOTE Algorithm for Improving Crop Growth Prediction When Climate Changes
College of Computer & Cyber Sciences, University of Prince Mugrin(穆格林王子大学计算机与网络科学学院);School of Computer Science, University College Dublin(都柏林大学学院计算机科学学院);Insight Centre for Data Analytics, University College Dublin(都柏林大学学院洞察数据分析中心);VistaMilk SFI Research Centre, University College Dublin(都柏林大学学院VistaMilk SFI研究中心)
School of Information Science and Technology, Beijing Forestry University(北京林业大学信息科学与技术学院);School of Computer Science & Technology, Beijing Institute of Technology(北京理工大学计算机科学与技术学院)
DOME: Improving Signal-to-Noise in Stochastic Gradient Descent via Sharp-Direction Subspace Filtering
International Laboratory on Learning Systems (ILLS), Montreal, Canada;Mila – Quebec AI Institute, Montreal, Canada;McGill University, Montreal, Canada;Inria, INSA Lyon, CITI, UR3720, 69621 Villeurbanne, France;CNRS, INSA Lyon, LIRIS, UMR 5205, 69621 Villeurbanne, France
Outlier Detection of Poisson-Distributed Targets Using a Seabed Sensor Network
Georgia Southern University(佐治亚南方大学);Virginia Polytechnic Institute and State University(弗吉尼亚理工大学和州大学);Johns Hopkins University Applied Physics Laboratory(约翰霍普金斯大学应用物理实验室)
ORSA: Outlier Robust Stacked Aggregation for Best- and Worst-Case Approximations of Ensemble Systems\
University of Stuttgart(斯图加特大学);Advantest Europe GmbH(爱德万测试欧洲有限公司)
近年来,集成学习在应用中的使用显著增长,这是由于计算能力的增加允许在合理的时间范围内训练大型集成。许多应用,例如,恶意软件检测、面部识别或财务决策使用有限的学习算法集合,并且以获得比任何其它单独的学习算法更好的预测性能的方式聚集它们。在半导体器件的硅后验证(Post-Silicon Validation for Semiconductor Devices,PSV)领域中,通常提供由各种器件组成的数据集,例如,不同生产线的芯片。在PSV中,任务是使用多个学习算法来近似数据的底层函数,每个学习算法都在特定于设备的子集上进行训练,而不是提高任意分类器在整个数据集上的性能。此外,期望是未知数量的子集描述显示非常不同特征的函数。相应的系综成员,被称为离群值,可以严重影响近似。我们的方法旨在找到一个合适的近似值,该近似值对离群值具有鲁棒性,并以适用于尽可能多的类型的方式代表最佳或最差情况。使用“soft-max”或“soft-min”函数代替最大值或最小值运算符。一个神经网络(NN)被训练成在两个阶段的过程中学习这个“软函数”。首先,我们选择代表最佳或最差情况的集合成员的子集。其次,我们结合这些成员,并定义一个权重,使用本地离群值因子(LOF)的属性,以增加非离群值的影响,并减少离群值。加权确保了对离群值的鲁棒性,并确保近似值适用于大多数类型。
Robust Mixture Learning when Outliers Overwhelm Small Groups
ETH Zurich(苏黎世联邦理工学院);TU Munich(慕尼黑工业大学);Lucerne School of Computer Science and Information Technology(卢塞恩计算机科学与信息技术学院);University of Copenhagen(哥本哈根大学)
Improving healthcare access management by predicting patient no-show behaviour
Southampton Business School, University of Southampton(南安普顿大学南安普顿商学院);Pontificia Universidad Javeriana(哈维里亚那天主教大学);The University of Western Ontario(西安大略大学);Mathematical Sciences, University of Southampton(南安普顿大学数学科学学院)
What Are the Odds? Improving the foundations of Statistical Model Checking
Lancaster University Leipzig(莱斯特大学莱比锡分校);Ruhr University Bochum(波鸿鲁尔大学);Institute of Science and Technology Austria(奥地利科学技术研究所);Dresden University of Technology(德累斯顿工业大学);Centre for Tactile Internet with Human-in-the-Loop (CeTI)(人机交互触觉互联网中心(CeTI))
ALDI++: Automatic and parameter-less discord and outlier detection for building energy load profiles
National University of Singapore(新加坡国立大学);Karlsruhe Institute of Technology(卡尔斯鲁厄理工学院);University of Texas at Arlington(德克萨斯大学阿灵顿分校)
数据驱动的建筑能耗预测是测量和验证、建筑基准测试以及建筑与电网交互过程中不可或缺的一部分。ASHRAE Great Energy Predictor III(GEPIII)机器学习竞赛使用广泛的仪表数据集众包最准确的机器学习工作流程,用于整个建筑物的能源预测。获胜解决方案的一个重要组成部分是预处理阶段,以消除异常的训练数据。当前的预处理方法集中于过滤统计阈值或需要训练数据和多个超参数的深度学习方法。最近的一种名为ALDI(自动加载配置文件不一致识别)的方法设法使用矩阵配置文件识别这些不一致,但该技术仍然需要用户定义的参数。我们开发ALDI++,一种基于以前的工作,绕过用户定义的参数,并利用不一致的相似性的方法。我们评估ALDI++对统计阈值,变分自动编码器,和原始ALDI作为基线分类不和谐和能源预测方案。我们的研究结果表明,虽然分类性能的改善,在原来的方法是边际,ALDI++有助于实现最佳的预测误差提高6%,比获胜的团队的方法少六倍的计算时间。
SmoothRot: Combining Channel-Wise Scaling and Rotation for Quantization-Friendly LLMs
Doctoral School of Applied Informatics and Applied Mathematics, Obuda University(应用信息学与应用数学博士学院,奥布达大学);John von Neumann Faculty of Informatics, Obuda University(约翰·冯·诺依曼信息学院,奥布达大学)