Download DFTS-P2miner (v2.2)

Tutorial, documentation, virtual machine

SITS and DFTS datasets

DFTS-P2miner: for extracting and ranking maximal reliable Grouped Frequent Sequential patterns (GFS-patterns) from Displacement Field Times Series (DFTS) and their confidence measures. A quantization procedure (equal frequency bucketting) and a pattern ranking module (swap randomization, normalized mutual information) are embodied. A running example is made available. It is based on the Mont Blanc DFTS used in the JSTARS article whose reference is given below.

Authors: Tuan Nguyen, Nicolas Méger (LISTIC), Christophe Rigotti (LIRIS), Catherine Pothier (LIRIS), Andreea Julea, Felicity Lodge, Quentin Chalabi, Aymeric Gaillard.

A pattern-based method for handling confidence measures while mining satellite displacement field time series. Application to Greenland ice sheet and Alpine glaciers. Tuan Nguyen, Nicolas Méger, Christophe Rigotti, Catherine Pothier, Emmanuel Trouvé, N. Gourmelen and J-L. Mugnier. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (JSTARS), 11 (11), pp.4390 - 4402. October 2018.

A pattern-based mining system for exploring Displacement Field Time Series. Tuan Nguyen, Nicolas Méger, Christophe Rigotti, Catherine Pothier, Noel Gourmelen, Emmanuel Trouvé. 19th IEEE International Conference on Data Mining (ICDM 2019) Demo, pp.1110-1113. Beijing, China, November 2019.