Output details

An unsupervised conditional random fields approach for clustering gene expression time series

<24> Published in one of the top journals in its field, this paper documents a novel unsupervised classifier, which can cluster gene expression time series without prior knowledge. The classifier is highly adaptable to various modalities of data, and has found application in conditional random fields (Foriselli, Bologna), analysing dynamic regulatory networks (Gitter, Microsoft Research) and clustering high-dimensional data (Chan, Melbourne). It is also being used for image segmentation and blind image classification e.g. by INTERPOL to add image/video retrieval functionality to its international child sexual exploitation database, said to be having “profound societal impact” (Sadeh, INTERPOL; Leary, Forensic Pathways Ltd).