tag:blogger.com,1999:blog-5246987755651065286.post4907672911311507941..comments2024-02-22T16:15:42.388-08:00Comments on cbloom rants: 09-12-10 - PPM vs CMcbloomhttp://www.blogger.com/profile/10714564834899413045noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-5246987755651065286.post-72914534784307331232010-09-13T00:15:12.217-07:002010-09-13T00:15:12.217-07:00Your "green" and "tree" progra...Your "green" and "tree" programs? Yeah, I downloaded them and like them very much, they're a good introduction to CM.<br /><br />Why is noone else doing bytewise mixing? Have you tried making the weighting more generic instead of hand-tweaked?<br /><br />Also, yeah I think it is a good way to see PPM as an optimization of CM. PPM is a particular type of simplified mixing in which you can stop doing any more work once you code from a context.cbloomhttps://www.blogger.com/profile/10714564834899413045noreply@blogger.comtag:blogger.com,1999:blog-5246987755651065286.post-39547847491569077492010-09-13T00:04:36.049-07:002010-09-13T00:04:36.049-07:001. Did you see this
http://encode.ru/threads/541...1. Did you see this <br /><a href="http://encode.ru/threads/541-Simple-bytewise-context-mixing-demo?p=10910&viewfull=1#post10910" rel="nofollow"><br />http://encode.ru/threads/541-Simple-bytewise-context-mixing-demo</a>?<br /><br />2. My opinion is that PPM is a speed optimization of CM - imagine taking a CM and caching the mixed freqs of lower orders in the context node.Shelwienhttps://www.blogger.com/profile/15845762957306674934noreply@blogger.com