Comparar métodos

	Detecção de Anomalias com Autoencoder Online	One-Class SVM
Área	Aprendizado de máquina	Aprendizado de máquina
Família	Machine learning	Machine learning
Ano de origem	2010s–present	1999–2001
Autor original	Various (online/incremental deep learning community)	Scholkopf, B., Platt, J. C., Smola, A. J., Williamson, R. C.
Tipo	Online unsupervised anomaly detection	Anomaly / novelty detection (unsupervised)
Fonte seminal	An, J. & Cho, S. (2015). Variational Autoencoder based Anomaly Detection using Reconstruction Probability. SNU Data Mining Center, 2015-2.	Scholkopf, B., Platt, J. C., Shawe-Taylor, J., Smola, A. J., & Williamson, R. C. (2001). Estimating the support of a high-dimensional distribution. Neural Computation, 13(7), 1443–1471.
Outros nomes	incremental autoencoder anomaly detection, streaming autoencoder anomaly detection, online AE anomaly detection, continual autoencoder anomaly detection	OCSVM, one-class support vector machine, novelty SVM, unsupervised SVM
Relacionados	5	3
Resumo	Online Autoencoder Anomaly Detection trains an autoencoder incrementally on a continuous data stream, flagging observations whose reconstruction error exceeds an adaptive threshold as anomalies. This approach combines the representational power of deep autoencoders with the incremental update capability of online learning, making it suitable for real-time or high-volume streaming scenarios where batch retraining is impractical.	One-class SVM is an unsupervised anomaly and novelty detection algorithm that learns a tight boundary around normal training data in a kernel-induced feature space, flagging new observations that fall outside that boundary as outliers. Introduced by Scholkopf et al. in 1999–2001, it extends the SVM framework to the single-class setting where no labelled anomalies are available.