Abstract | ||
---|---|---|
In the paper we present a new framework for dealing with probabilistic graphical models. Our approach relies on the recently proposed Tensor Train format (TT-format) of a tensor that while being compact allows for efficient application of linear algebra operations. We present a way to convert the energy of a Markov random field to the TT-format and show how one can exploit the properties of the TT-format to attack the tasks of the partition function estimation and the MAP-inference. We provide theoretical guarantees on the accuracy of the proposed algorithm for estimating the partition function and compare our methods against several state-of-the-art algorithms. |
Year | Venue | Field |
---|---|---|
2014 | ICML | Linear algebra,Tensor,Markov random field,Partition function (statistical mechanics),Computer science,Exploit,Theoretical computer science,Artificial intelligence,Tensor train,Graphical model,Machine learning |
DocType | Volume | Issue |
Conference | 32 | 1 |
Citations | PageRank | References |
5 | 0.56 | 17 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Alexander Novikov | 1 | 98 | 7.62 |
Anton Rodomanov | 2 | 5 | 0.56 |
A. Osokin | 3 | 430 | 19.01 |
Dmitry Vetrov | 4 | 263 | 21.56 |