Noiseless Independent Factor Analysis with Mixing Constraints in a Semi-supervised Framework. Application to Railway Device Fault Diagnosis
Résumé
In Independent Factor Analysis (IFA), latent components (or sources) are only recovered from their linear observed mixtures. Both the mixing process and the sources densities (that are assumed to be generated according to mixtures of gaussians) are learned from observed data. This paper investigates the possibility of estimating the IFA model when two prior knowledge are incorporated : constraints on the mixing process and partial knowledge on the cluster membership of some examples. Semi-supervised or partially supervised learning frameworks can thus be handled. These two proposals have been initially motivated by a real-world application that concerns a fault diagnosis of a railway device. Results on this application are provided to demonstrate its ability to enhance estimation accuracy and remove indeterminacy commonly encountered in unsupervised IFA such as the sources permutations. \keywords{Independent Factor Analysis, mixing constraints, semi-supervised learning, diagnosis, railway device}
Domaines
Machine Learning [stat.ML]Origine | Fichiers produits par l'(les) auteur(s) |
---|
Loading...