Inferring Latent Structure From Mixed Real and Categorical Relational Data
dc.contributor.author | Salazar, E | |
dc.contributor.author | Cain, MS | |
dc.contributor.author | Darling, EF | |
dc.contributor.author | Mitroff, SR | |
dc.contributor.author | Carin, L | |
dc.date.accessioned | 2014-07-22T16:18:52Z | |
dc.description.abstract | We consider analysis of relational data (a matrix), in which the rows correspond to subjects (e.g., people) and the columns correspond to attributes. The elements of the matrix may be a mix of real and categorical. Each subject and attribute is characterized by a latent binary feature vector, and an inferred matrix maps each row-column pair of binary feature vectors to an observed matrix element. The latent binary features of the rows are modeled via a multivariate Gaussian distribution with low-rank covariance matrix, and the Gaussian random variables are mapped to latent binary features via a probit link. The same type construction is applied jointly to the columns. The model infers latent, low-dimensional binary features associated with each row and each column, as well correlation structure between all rows and between all columns. | |
dc.identifier | ||
dc.identifier.uri | ||
dc.publisher | icml.cc / Omnipress | |
dc.subject | cs.LG | |
dc.subject | cs.LG | |
dc.subject | stat.ML | |
dc.title | Inferring Latent Structure From Mixed Real and Categorical Relational Data | |
dc.type | Journal article | |
pubs.author-url | ||
pubs.notes | Appears in Proceedings of the 29th International Conference on Machine Learning (ICML 2012) | |
pubs.organisational-group | Center for Cognitive Neuroscience | |
pubs.organisational-group | Duke | |
pubs.organisational-group | Duke Institute for Brain Sciences | |
pubs.organisational-group | Electrical and Computer Engineering | |
pubs.organisational-group | Faculty | |
pubs.organisational-group | Institutes and Provost's Academic Units | |
pubs.organisational-group | Pratt School of Engineering | |
pubs.organisational-group | Psychology and Neuroscience | |
pubs.organisational-group | Trinity College of Arts & Sciences | |
pubs.organisational-group | University Institutes and Centers |
Files
Original bundle
- Name:
- 1206.6469v1.pdf
- Size:
- 1.02 MB
- Format:
- Adobe Portable Document Format
- Description:
- Published version