ssaneine at gmail.com
Fri Nov 26 13:42:42 PST 2010
Is this in R? Are you using power iteration? I'm considering both these
approaches for memory footprint and code simplicity reasons, respectively
(my laptop is a little long in the tooth).
Also, what's "quasi" about your quasi-PageRank feature?
On Thu, Nov 25, 2010 at 4:11 PM, Jared Dunne <jareddunne at gmail.com> wrote:
> I've implemented an iterative approach to quasi-PageRank feature generation
> for all nodes. Each iteration across all nodes is under a minute each
> time. I still need to fix some things with my approach and possibly get
> some help sanity-checking the correctness of my algorithm. I'll post more
> later, but I just wanted to let everyone know that I'm making good progress
> on the PageRank feature generation, so that we aren't duplicating our
> efforts. Gonna pick this back up later this weekend. Enjoy your feasts.
> On Wed, Nov 24, 2010 at 6:23 PM, <mike at mindmech.com> wrote:
>> I've shared SocialPageRank<https://docs.google.com/present/edit?id=0AbXOdbbcuXxAZGNidmNnNjJfMGRkNmg4OGR2&hl=en&invite=CO_7wcEI>
>> Message from mike at mindmech.com:
>> Here's the presentation I'll be giving tonight on Random Walks, PageRank, and Social Net Analysis.
>> Click to open:
>> - SocialPageRank<https://docs.google.com/present/edit?id=0AbXOdbbcuXxAZGNidmNnNjJfMGRkNmg4OGR2&hl=en&invite=CO_7wcEI>
>> Google Docs makes it easy to create, store and share online documents,
>> spreadsheets and presentations.
>> [image: Logo for Google Docs] <http://docs.google.com>
>> ml mailing list
>> ml at lists.noisebridge.net
> ml mailing list
> ml at lists.noisebridge.net
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the ml