Estimation of Parsimonious Covariance Models for Gaussian Matrix Valued Random Variables for Multi-Dimensional Spectroscopic Data

Asmita Poddar (1) , Serge Iovleff (2) , Florent Latimier (2)

1 NUS - National University of Singapore
2 MODAL - MOdel for Data Analysis and Learning

Satellite remote sensing makes it possible to observe landscapes on large spatial scales. The Sentinel-1 and Sentinel-2 satellites currently provide full coverage of the national territory of France every 5 days. Due to the orbit of the satellites, coupled with the presence of clouds, the sampling of the pixels are temporally irregular. The project aims to develop, study and implement supervised and unsupervised classification methods when the data are of different natures (heterogeneous) and have missing and/or aberrant data. The methods implemented are developed to process satellite and aerial data for ecology and cartography.

Domaines: Applications [stat.AP]; Statistiques [math.ST]

Asmita Poddar, Serge Iovleff, Florent Latimier. Estimation of Parsimonious Covariance Models for Gaussian Matrix Valued Random Variables for Multi-Dimensional Spectroscopic Data. WiML 2018 - 13th Women in Machine Learning workshop, Dec 2018, Montreal, Canada. 2018. ⟨hal-01954769⟩