hoya012 / awesome-anomaly-detection Goto Github PK

Hi there,

Looks like SDM17 paper has an incorrect author/pdf info.

Chen, J., Sathe, S., Aggarwal, C. and Turaga, D., 2017, June. Outlier detection with autoencoder ensembles. In Proceedings of the 2017 SIAM International Conference on Data Mining (pp. 90-98). Society for Industrial and Applied Mathematics.

pdf: http://saketsathe.net/downloads/autoencode.pdf

Thanks,
Yue

Hi,

This paper has been accepted in CVPR2021. I kindly ask you to add it to your nice repo.
"Multiresolution Knowledge Distillation for Anomaly Detection" : https://arxiv.org/pdf/2011.11108.pdf

Thanks.

"Anomaly Detection in Multivariate Non-stationary Time Series for Automatic DBMS Diagnosis"
Is a paper introduced in ICMLA 2017 (International Conference of Machine Learning and Applications."
It detect abnormal changes by auto encoder and statistical process control method.
Also introduce an application capture root causes with time series similarity measure, DTW.

Hi @hoya012, thank you for the super awesome repository!

First, thank you for mentioning our paper "CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instance" (NeurIPS 2020). However, I want to clarify that our main category is one-class (anomaly) classification, while we also provide an extension for (multi-class) OOD detection.

Can you move our paper to the one-class classification (or both one-class and OOD) category?

Thank you for your kind response and valuable efforts!

Hi @hoya012
I want to add some more papers for time series anomaly detection.

However, as you know well, there are so many papers in arXiv...
So I just wonder is there any critera to select papers here, like only papers that you read or something like that.

If there is any, please let me know.
I'll send a pr with paper references under your criteria.

Thanks for creating such a nice repo!