Filters








8,924 Hits in 3.9 sec

Localized Multiple Kernel Learning for Anomaly Detection: One-class Classification [article]

Chandan Gautam, Ramesh Balaji, K Sudharsan, Aruna Tiwari, Kapil Ahuja
2018 arXiv   pre-print
In this paper, we present a multiple kernel learning approach for the One-class Classification (OCC) task and employ it for anomaly detection.  ...  This paper proposes a Localized Multiple Kernel learning approach for Anomaly Detection (LMKAD) using OCC, where the weight for each kernel is assigned locally.  ...  Localized Multiple Kernel Anomaly Detection In this section, we propose Localized Multiple Kernel Anomaly Detection (LMKAD).  ... 
arXiv:1805.07892v4 fatcat:m372ynocw5gebf24viodbna4ly

Self-Trained One-class Classification for Unsupervised Anomaly Detection [article]

Jinsung Yoon, Kihyuk Sohn, Chun-Liang Li, Sercan O. Arik, Chen-Yu Lee, Tomas Pfister
2021 arXiv   pre-print
For example, with a 10% anomaly ratio on CIFAR-10 data, the proposed method outperforms state-of-the-art one-class classification method by 6.3 AUC and 12.5 average precision.  ...  In experiments, we show the efficacy of our method for unsupervised anomaly detection on benchmarks from image and tabular data domains.  ...  learning strategies of one-class classification for unsupervised anomaly detection.  ... 
arXiv:2106.06115v1 fatcat:nt5m7se5pvcn5lckwz6edncgkq

Classifier-Adjusted Density Estimation for Anomaly Detection and One-Class Classification [chapter]

Lisa Friedland, Amanda Gentzel, David Jensen
2014 Proceedings of the 2014 SIAM International Conference on Data Mining  
Density estimation methods are often regarded as unsuitable for anomaly detection in high-dimensional data due to the difficulty of estimating multivariate probability distributions.  ...  Instead, the scores from popular distance-and localdensity-based methods, such as local outlier factor (LOF), are used as surrogates for probability densities.  ...  Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation here on.  ... 
doi:10.1137/1.9781611973440.67 dblp:conf/sdm/FriedlandGJ14 fatcat:yx6btwcedjbexh4qxtk3tv6ulu

Learning and Evaluating Representations for Deep One-class Classification [article]

Kihyuk Sohn, Chun-Liang Li, Jinsung Yoon, Minho Jin, Tomas Pfister
2021 arXiv   pre-print
In experiments, we demonstrate state-of-the-art performance on visual domain one-class classification benchmarks, including novelty and anomaly detection.  ...  We present a two-stage framework for deep one-class classification.  ...  Classification-based anomaly detection for general data.  ... 
arXiv:2011.02578v2 fatcat:y2m2foa7i5ayrmabwxbnjjqrn4

DROCC: Deep Robust One-Class Classification [article]

Sachin Goyal, Aditi Raghunathan, Moksh Jain, Harsha Vardhan Simhadri, Prateek Jain
2020 arXiv   pre-print
Classical approaches for one-class problems such as one-class SVM and isolation forest require careful feature engineering when applied to structured domains like images.  ...  and ImageNet), audio, and time-series, offering up to 20% increase in accuracy over the state-of-the-art in anomaly detection.  ...  AR was funded by an Open Philanthropy AI Fellowship and Google PhD Fellowship in Machine Learning.  ... 
arXiv:2002.12718v2 fatcat:3vxztrzyrvfz3k7v7cjr2fsyqa

One-Class Classification: A Survey [article]

Pramuditha Perera, Poojan Oza, Vishal M. Patel
2021 arXiv   pre-print
One-Class Classification (OCC) is a special case of multi-class classification, where data observed during training is from a single positive class.  ...  In this article, we provide a survey of classical statistical and recent deep learning-based OCC methods for visual recognition.  ...  In this survey, we make no distinction between OCC and one class novelty detection. Outlier detection (unsupervised anomaly detection).  ... 
arXiv:2101.03064v1 fatcat:gcy4c5sjgrhilev54qttwpmdoe

Anomaly Detection Using Signal Segmentation and One-Class Classification in Diffusion Process of Semiconductor Manufacturing

Kyuchang Chang, Youngji Yoo, Jun-Geol Baek
2021 Sensors  
Phase I has three steps: signal segmentation, feature extraction based on local outlier factors (LOF), and one-class classification (OCC) modeling using the isolation forest (iF) algorithm.  ...  Much of the data gathered during this process is time series data for fault detection and classification (FDC) in real time.  ...  The entire framework consists of two phases, one for training and the other for testing. Each phase comprises three steps for anomaly detection.  ... 
doi:10.3390/s21113880 fatcat:s7wwm6nqdng2tpqpotw7yumcyu

One-Class Graph Neural Networks for Anomaly Detection in Attributed Networks [article]

Xuhong Wang, Baihong Jin, Ying Du, Ping Cui, Yupu Yang
2020 arXiv   pre-print
In this work, we propose One Class Graph Neural Network (OCGNN), a one-class classification framework for graph anomaly detection.  ...  As one of the dominant anomaly detection algorithms, One Class Support Vector Machine has been widely used to detect outliers.  ...  Conclusion In this paper, we proposed One-Class Graph Neural Network (OCGNN), a one-class classification framework for graph anomaly detection.  ... 
arXiv:2002.09594v2 fatcat:pdryiqoqs5ax3kmccjozz7mesm

One-Class Classification Based on Extreme Learning and Geometric Class Information

Alexandros Iosifidis, Vasileios Mygdalis, Anastasios Tefas, Ioannis Pitas
2016 Neural Processing Letters  
In this paper, we propose an Extreme Learning Machine-based one-class classification method that exploits geometric class information.  ...  proposed one-class classifiers.  ...  One-class classification (sometimes also called novelty detection, outlier detection, or anomaly detection) refers to the classification case where the available training data come from only one class,  ... 
doi:10.1007/s11063-016-9541-y fatcat:xvbnyhgt5vckzkxifm6zissxpq

Learning Deep Features for One-Class Classification [article]

Pramuditha Perera, Vishal M. Patel
2019 arXiv   pre-print
We propose a deep learning-based solution for the problem of feature learning in one-class classification.  ...  Extensive experiments on publicly available anomaly detection, novelty detection and mobile active authentication datasets show that the proposed Deep One-Class (DOC) classification method achieves significant  ...  A detailed description of various anomaly detection methods can be found in [6] . Novelty detection based on one-class learning has also received a significant attention in recent years.  ... 
arXiv:1801.05365v2 fatcat:pdx5hycy4jdizopsavf3ihjjm4

IoTDS: A One-Class Classification Approach to Detect Botnets in Internet of Things Devices

Vitor Hugo Bezerra, Victor Guilherme Turrisi da Costa, Sylvio Barbon Junior, Rodrigo Sanches Miani, Bruno Bogaz Zarpelão
2019 Sensors  
It relies on one-class classifiers, which model only the legitimate device behaviour for further detection of deviations, avoiding the manual labelling process.  ...  Four one-class algorithms (Elliptic Envelope, Isolation Forest, Local Outlier Factor, and One-class Support Vector Machine) were evaluated.  ...  A one-class classifier performs the classification.  ... 
doi:10.3390/s19143188 fatcat:y2gga4mgtfbxnknvsm3asnauge

Optimised one-class classification performance [article]

Oliver Urs Lenz, Daniel Peralta, Chris Cornelis
2021 arXiv   pre-print
We provide a thorough treatment of one-class classification with hyperparameter optimisation for five data descriptors: Support Vector Machine (SVM), Nearest Neighbour Distance (NND), Localised Nearest  ...  Neighbour Distance (LNND), Local Outlier Factor (LOF) and Average Localised Proximity (ALP).  ...  Introduction The goal of one-class classification (Tax (2001) , also known as novelty, semisupervised outlier or semi-supervised anomaly detection), is to form, on the basis of a representative sample  ... 
arXiv:2102.02618v2 fatcat:juhsnivutbgvlkyi6nucoydxzy

Relational One-Class Classification: A Non-Parametric Approach

Tushar Khot, Sriraam Natarajan, Jude Shavlik
2014 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
One-class classification approaches have been proposed in the literature to learn classifiers from examples of only one class.  ...  We propose a non-parametric relational one-class classification approach based on first-order trees.  ...  Anomaly/outlier detection methods can also be viewed as a one-class classification approach.  ... 
doi:10.1609/aaai.v28i1.9072 fatcat:z5vy7cbnlvhqtoa6qln5na3s6y

FROCC: Fast Random projection-based One-Class Classification [article]

Arindam Bhattacharya and Sumanth Varambally and Amitabha Bagchi and Srikanta Bedathur
2021 arXiv   pre-print
We present Fast Random projection-based One-Class Classification (FROCC), an extremely efficient method for one-class classification.  ...  FROCC achieves up to 3.1 percent points better ROC, with 1.2--67.8x speedup in training and test times over a range of state-of-the-art benchmarks including the SVM and the deep learning based models for  ...  Introduction One-class classification (OCC) has attracted a lot of attention over the years under various names such as novelty detection [30] , anomaly or outlier detection [34, 29] , concept learning  ... 
arXiv:2011.14317v3 fatcat:f6ccpqnxnfcx7d6rkkgnrylm2e

One-Class Classification by Ensembles of Regression models – a detailed study [article]

Amir Ahmad, Srikanth Bezawada
2020 arXiv   pre-print
One-class classification (OCC) deals with the classification problem in which the training data has data points belonging only to target class.  ...  In this paper, we study a one-class classification algorithm, One-Class Classification by Ensembles of Regression models (OCCER), that uses regression methods to address OCC problems.  ...  [7] propose the use of support vector machines for one class classification.  ... 
arXiv:1912.11475v3 fatcat:lba6tlyomjdfvaduzgvdxmohn4
« Previous Showing results 1 — 15 out of 8,924 results