Conditional random fields

Search Options	close
Search the following clips: All Clips

news science politics food economy art
technology health internet religion psychology

Sign Up Install Learn More Login


	Conditional random fields

fschilder

11-21-2006 11:05 PM

165 views

tags:

crf, machine learning, hmm

fschilder says:

Definition of CRFs, good introduction to CRFs

Add a Comment

Login to Comment. Not a member yet? Sign up

Today's Top Clips

visit the Top Clips page

View the Top Clips from November 21, 2006

Embed This Clip In Your Site...

<div style="margin: 12px 0px; font-family: arial; color: #333333; background: #ffffff; border: solid 4px #e5e5e5; width: 100%; clear: left;"><div class="CM_CTB_Content_Wrap" style="margin: 0px; padding: 0px;background-color: #ffffff;"><div style="border-bottom: solid 1px #dcdcdc; white-space: nowrap; margin-bottom: 8px; background-color: #eeeeee ;background-image: url(http://clipmarks.com/images/source-bg.gif); background-repeat: repeat-x; height: 24px; line-height: 24px; vertical-align: middle; padding-bottom: 4px; color: #666666; font-size: 10px;" ><a href="http://clipmarks.com/clip-to-blog/" title="see clips that are hot right now"><img src="http://content.clipmarks.com/blog_embed/5626f50e-2e07-482f-830e-f19059c9c28a/7C3224AC-0E1C-4797-B738-E64789CF9AF3/" alt="" width="19" height="19" border="0" style="vertical-align: middle; margin: 0px 4px; display: inline; border: none; float:none;" /></a>clipped from <a title="http://www.inference.phy.cam.ac.uk/hmw26/crf/" href="http://www.inference.phy.cam.ac.uk/hmw26/crf/" style="font-size: 11px;">www.inference.phy.cam.ac.uk</a></div><blockquote style="text-align: left; padding: 0px 8px; margin: 4px 0px 8px 0px; background: transparent; border: none;" cite="http://www.inference.phy.cam.ac.uk/hmw26/crf/"><P>Conditional random fields (CRFs) are a probabilistic framework for
labeling and segmenting structured data, such as sequences, trees and
lattices. The underlying idea is that of defining a conditional
probability distribution over label sequences given a particular
observation sequence, rather than a joint distribution over both label
and observation sequences. The primary advantage of CRFs over hidden
Markov models is their conditional nature, resulting in the relaxation
of the independence assumptions required by HMMs in order to ensure
tractable inference. Additionally, CRFs avoid the label bias problem,
a weakness exhibited by maximum entropy Markov models (MEMMs) and
other conditional Markov models based on directed graphical
models. CRFs outperform both MEMMs and HMMs on a number of real-world
tasks in many fields, including bioinformatics, computational
linguistics and speech recognition.</P></blockquote></div><div style="margin: 0px 6px 6px 4px;"><table style="font-size: 11px;border-spacing: 0px;padding: 0px;" cellpadding="0" cellspacing="0" width="100%"><tr><td style="background:transparent;border-width:0px;padding:0px;">&nbsp;</td><td align="right" style="background:transparent;border-width:0px;padding:0px;width:107px" width="107"><a href="http://clipmarks.com/share/7C3224AC-0E1C-4797-B738-E64789CF9AF3/blog/" title="blog or email this clip"><img src="http://content9.clipmarks.com/images/c2b-foot.png" border="0" alt="blog it" width="107" height="17" style="border-width:0px;padding:0px;margin:0px;" /></a></td></tr></table></div></div>

New from the makers of Clipmarks: Amplify.com - Don't just share the news...Amplify it!

Clipmarks: Home; New Clips; Top Clips; Dashboard

Popular Topics: News; Life; Science; Technology; Entertainment

Get Started: Sign Up; Install Clipping Tool; How Clipping Works; Clip-to-Blog™; ClipSearch

Tools and Resources: FAQ; ClipWeek; Top Clippers; Top Tags; Site Map

About Clipmarks: About Us; Contact; Copyright; Privacy; EULA

Search Options

Today's Top Clips