We can reduce memory complexity of hmm inference to

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: h Time/space tradeoffs We can pay more (in terms of time) to recover some (in terms of memory). Prof. Jeff Bilmes EE596A/Winter 2013/DGMs – Lecture 9 - Feb 6th, 2013 page 9-32 (of 180) k-best without the penalty Island Summary Scratch Time/space tradeoffs We can pay more (in terms of time) to recover some (in terms of memory). We can reduce memory complexity of HMM inference to O(N log T ) while the compute complexity increase to O(N 2 T log T ). Prof. Jeff Bilmes EE596A/Winter 2013/DGMs – Lecture 9 - Feb 6th, 2013 page 9-32 (of 180) k-best without the penalty Island Summary Scratch Time/space tradeoffs We can pay more (in terms of time) to recover some (in terms of memory). We can reduce memory complexity of HMM inference to O(N log T ) while the compute complexity increase to O(N 2 T log T ). That is, additional factor of computation log T modest, while the reduction factor in memory is large T / log T . Prof. Jeff Bilmes EE596A/Winter 2013/DGMs – Lecture 9 - Feb 6th, 2013 page 9-32 (of 180) k-best without the penalty Island Summary Scratch Time/space tradeoffs We can pay more (in terms of time) to recover some (in terms of memory). We can reduce memory complexity of HMM inference to O(...
View Full Document

Ask a homework question - tutors are online