Publication · 2001-2007

Pickett - temporal pattern learning papers (ICML 2002, AAAI 2001, AAAI 2007)

Archived record of the Pickett papers on discovering temporal patterns in multivariate time series, presented at ICML and AAAI.

This line of work addressed unsupervised discovery of recurring patterns in multivariate sensor streams - finding the events in the signal before anyone has labelled them.

It is the direct ancestor of condition-monitoring analytics on inspection robots: a quadruped walking a fixed route produces exactly this kind of repeated multivariate series, and the useful output is the deviation.

This page preserves a historic coral-lab.org address. The original files are held offline; if you need them for citation or attribution, contact Coral Labs.