SOTAVerified

One-Class Classification

One-class classification (OCC) algorithms serve a crucial role in scenarios where the negative class is either absent, poorly sampled, or not well defined. This unique situation presents a challenge for building effective classifiers, as they must delineate the class boundary solely based on knowledge of the positive class. OCC has found application in various research domains, including outlier/novelty detection and concept learning.

In the context of anomaly detection, OCC models are trained exclusively on "normal" data and are subsequently tasked with identifying anomalous patterns during inference.

A one-class classifier aims at capturing characteristics of training instances, in order to be able to distinguish between them and potential outliers to appear.

— Page 139, Learning from Imbalanced Data Sets, 2018.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–10 of 227 papers

TitleStatusHype
SD-MAD: Sign-Driven Few-shot Multi-Anomaly Detection in Medical Images—0
Adversarial Subspace Generation for Outlier Detection in High-Dimensional DataCode0
RoCA: Robust Contrastive One-class Time Series Anomaly Detection with Contaminated DataCode0
From Pixels to Trajectory: Universal Adversarial Example Detection via Temporal Imprints—0
FedDyMem: Efficient Federated Learning with Dynamic Memory and Memory-Reduce for Unsupervised Image Anomaly Detection—0
Anomaly Detection in Smart Power Grids with Graph-Regularized MS-SVDD: a Multimodal Subspace Learning Approach—0
One Class Restricted Kernel MachinesCode0
Score Combining for Contrastive OOD Detection—0
Teacher Encoder-Student Decoder Denoising Guided Segmentation Network for Anomaly Detection—0
On the Adversarial Robustness of Benjamini Hochberg—0
Show:102550
← PrevPage 1 of 23Next →

No leaderboard results yet.