Infinite dynamic bayesian networks
We present the infinite dynamic Bayesian network model (iDBN), a nonparametric, factored state-space model that generalizes dynamic Bayesian networks (DBNs). The iDBN can infer every aspect of a DBN: the number of hidden factors, the number of values each factor can take, and (arbitrarily complex) c...
| Main Authors: | , , , |
|---|---|
| Andre forfattere: | |
| Format: | Article |
| Sprog: | en_US |
| Udgivet: |
International Machine Learning Society
2012
|
| Online adgang: | http://hdl.handle.net/1721.1/70126 https://orcid.org/0000-0002-1925-2035 https://orcid.org/0000-0002-8293-0492 |
| Summary: | We present the infinite dynamic Bayesian network model (iDBN), a nonparametric, factored state-space model that generalizes dynamic Bayesian networks (DBNs). The iDBN can infer every aspect of a DBN: the number of hidden factors, the number of values each factor can take, and (arbitrarily complex) connections and conditionals between factors and observations. In this way, the iDBN generalizes other nonparametric state space models, which until now generally
focused on binary hidden nodes and more restricted connection structures. We show how
this new prior allows us to find interesting structure in benchmark tests and on two realworld datasets involving weather data and neural information flow networks. |
|---|