Message Passing In Centralized Database
Add to your list(s)
Download to your calendar using vCal
If you have a question about this talk, please contact Zoubin Ghahramani.
short talk
This talk discusses the results found during an internship research project applying Bayesian inference on centralized database architecture.
Probabilistic graphical models are rapidly finding their way into commercial products. In many of these environments, the data on which the graphical models operate live is stored in structured relational databases. In this project we investigate whether we can implement inference in graphical models within the database.
In this talk I will discuss how we deployed the declarative SQL language and built an expectation propagation based algorithm for the TrueSkill model in a relational database. I will discuss the performance characteristics of such an architecture and compare them to a Hadoop style map-reduce implementation of message passing.
This talk is part of the Machine Learning @ CUED series.
This talk is included in these lists:
Note that ex-directory lists are not shown.
|