To refer to this page use:
|Abstract:||Latent feature models are widely used to decompose data into a small number of components. Bayesian nonparametric variants of these models, which use the Indian buffet process (IBP) as a prior over latent features, allow the number of features to be determined from the data. We present a generalization of the IBP, the distance dependent Indian buffet process (dd-IBP), for modeling non-exchangeable data. It relies on distances defined between data points, biasing nearby data to share more features. The choice of distance measure allows for many kinds of dependencies, including temporal and spatial. Further, the original IBP is a special case of the dd-IBP. We develop the dd-IBP and theoretically characterize its feature-sharing properties. We derive a Markov chain Monte Carlo sampler for a linear Gaussian model with a dd-IBP prior and study its performance on real-world non-exchangeable data.|
|Citation:||Gershman, Samuel J., Peter I. Frazier, and David M. Blei. "Distance Dependent Infinite Latent Feature Models." IEEE Transactions on Pattern Analysis and Machine Intelligence 37, no. 2 (2014): pp. 334-345. doi:10.1109/TPAMI.2014.2321387|
|Pages:||334 - 345|
|Type of Material:||Journal Article|
|Journal/Proceeding Title:||IEEE Transactions on Pattern Analysis and Machine Intelligence|
Items in OAR@Princeton are protected by copyright, with all rights reserved, unless otherwise indicated.