Anomaly detection

I don’t define what is normal, but I know it when I see it

Working out when something you are seeing is not something you should have been expecting, in some sense which will not be made rigorous here, not yet.

Nuit Blanche has a roundup.

Special case: “trend detection”. e.g. Gnip Trend detection Nikolov (2012) looks fun, and comes with a cute explanation of how you might do this nonparametrically.

Henrickson, Scott, Jeff Kolb, Brian Lehmann, and Joshua Montague. 2015. “Trend Detection in Social Data.” Twitter. https://github.com/jeffakolb/Gnip-Trend-Detection.

Kandanaarachchi, Sevvandi, Mario A. Munoz, Rob J. Hyndman, and Kate Smith-Miles. 2018. “On Normalization and Algorithm Selection for Unsupervised Outlier Detection.” 16/18. Monash Econometrics and Business Statistics Working Papers. Monash University, Department of Econometrics and Business Statistics. https://www.monash.edu/business/ebs/research/publications/ebs/wp16-2018.pdf.

Lopez, Jose A., Octavia Camps, and Mario Sznaier. 2015. “Robust Anomaly Detection Using Semidefinite Programming,” April. http://arxiv.org/abs/1504.00905.

Nikolov, Stanislav. 2012. “Trend or No Trend : A Novel Nonparametric Method for Classifying Time Series.” Thesis, Massachusetts Institute of Technology. http://dspace.mit.edu/handle/1721.1/85399.

She, Yiyuan, and Art B. Owen. 2010. “Outlier Detection Using Nonconvex Penalized Regression.” http://statweb.stanford.edu/~owen/reports/theta-ipod.pdf.