Sammanfattning
We consider independent component analysis of binary data. While fundamental in practice, this case has been much less developed than ICA for continuous data. We start by assuming a linear mixing model in a continuous-valued latent space, followed by a binary observation model. Importantly, we assume that the sources are non-stationary; this is necessary since any non-Gaussianity would essentially be destroyed by the binarization. Interestingly, the model allows for closed-form likelihood by employing the cumulative distribution function of the multivariate Gaussian distribution. In stark contrast to the continuous-valued case, we prove non-identifiability of the model with few observed variables; our empirical results imply identifiability when the number of observed variables is higher. We present a practical method for binary ICA that uses only pairwise marginals, which are faster to compute than the full multivariate likelihood. Experiments give insight into the requirements for the number of observed variables, segments, and latent sources that allow the model to be estimated.
Originalspråk | engelska |
---|---|
Titel på värdpublikation | The 38th Conference on Uncertainty in Artificial Intelligence |
Antal sidor | 11 |
Förlag | The Association for Uncertainty in Artificial Intelligence |
Utgivningsdatum | 2022 |
Sidor | 874-884 |
ISBN (elektroniskt) | 978-1-7138-6329-8 |
Status | Publicerad - 2022 |
MoE-publikationstyp | A4 Artikel i en konferenspublikation |
Evenemang | Conference on Uncertainty in Artificial Intelligence - Eindhoven University of Technology, Eindhoven, Holland Varaktighet: 2 aug. 2022 → 4 aug. 2022 Konferensnummer: 38 https://www.auai.org/uai2022/ |
Bibliografisk information
Publisher Copyright:© 2022 Proceedings of the 38th Conference on Uncertainty in Artificial Intelligence, UAI 2022. All right reserved.
Vetenskapsgrenar
- 113 Data- och informationsvetenskap