Enhanced NMF initialization using a physical model for pollution source apportionment

Marc Plouvin (1) , Abdelhakim Limem (1) , Matthieu Puigt (1) , Gilles Delmaire (1) , Gilles Roussel (1) , Dominique Courcot (2)

1 LISIC - Laboratoire d'Informatique Signal et Image de la Côte d'Opale
2 UCEIV - Unité de Chimie Environnementale et Interactions sur le Vivant

In a previous work, we proposed an informed Non-negative Matrix Factorization (NMF) with a specific parametrization which involves constraints about some known components of the factorization. In this paper we extend the above work by adding some information provided by a physical dispersion model. In particular, we derive a special structure of one of the factorizing matrices, which provides a better initialization of the NMF procedure. Experiments on simulated mixtures of particulate matter sources show that our new approach outperforms both our previous one and the state-of-the-art NMF methods.

Domaines: Traitement du signal et de l'image [eess.SP]

Marc Plouvin, Abdelhakim Limem, Matthieu Puigt, Gilles Delmaire, Gilles Roussel, et al.. Enhanced NMF initialization using a physical model for pollution source apportionment. 22nd European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2014), Apr 2014, Brugge, Belgium. pp.261-266. ⟨hal-01367324⟩