Applying separative non-negative matrix factorization to extra-financial data

P Fogel (1) , C Geissler (1) , P Cotte (1) , G Luta (2)

1 Advestis
2 GU - Georgetown University [Washington]

We present here an original application of the non-negative matrix factorization (NMF) method, for the case of extra-financial data. These data are subject to high correlations between co-variables, as well as between observations. NMF provides a much more relevant clustering of co-variables and observations than a simple principal component analysis (PCA). In addition, we show that an initial data separation step before applying NMF further improves the quality of the clustering.

Mots clés: Machine Learning; Dimension reduction; Clustering; Interpretability; Features Engineering; ESG data; Machine Learning

Domaines: Machine Learning [stat.ML]; Intelligence artificielle [cs.AI]; Finance quantitative [q-fin.CP]

P Fogel, C Geissler, P Cotte, G Luta. Applying separative non-negative matrix factorization to extra-financial data. Bankers Markets & Investors : an academic & professional review, In press. ⟨hal-03141876v2⟩