MML Probabilistic Principal Component Analysis
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , |
|---|---|
| Format: | Preprint |
| Publié: |
2022
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866912888844713984 |
|---|---|
| author | Makalic, Enes Schmidt, Daniel F. |
| author_facet | Makalic, Enes Schmidt, Daniel F. |
| contents | Principal component analysis (PCA) is perhaps the most widely used method for data dimensionality reduction. A key question in PCA is deciding how many factors to retain. This manuscript describes a new approach to automatically selecting the number of principal components based on the Bayesian minimum message length method of inductive inference. We derive a new estimate of the isotropic residual variance and demonstrate that it improves on the usual maximum likelihood approach. We also discuss extending this approach to finite mixture models of principal component analyzers. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2209_14559 |
| institution | arXiv |
| publishDate | 2022 |
| record_format | arxiv |
| spellingShingle | MML Probabilistic Principal Component Analysis Makalic, Enes Schmidt, Daniel F. Methodology Principal component analysis (PCA) is perhaps the most widely used method for data dimensionality reduction. A key question in PCA is deciding how many factors to retain. This manuscript describes a new approach to automatically selecting the number of principal components based on the Bayesian minimum message length method of inductive inference. We derive a new estimate of the isotropic residual variance and demonstrate that it improves on the usual maximum likelihood approach. We also discuss extending this approach to finite mixture models of principal component analyzers. |
| title | MML Probabilistic Principal Component Analysis |
| topic | Methodology |
| url | https://arxiv.org/abs/2209.14559 |