FAQ
We decide to start with reducing the complexity of the process for
now, but it seems that our recommendation system and n-clustering
algoritms using just commands from Redis sets are enough, but it seems
we will have a lot of problems with scalability on this, but we
already have some engineers working with Mahout.

Our service is exactly a mix of all features from the social networks
(+tumblr) that you have talked, but we will just accept content that
meets our guidelines that will be anything related to knowledge , of
course, programming will be the most popular =)

Thanks,
Vinicius Melo

On Sun, Jul 29, 2012 at 11:28 PM, M. Edward (Ed) Borasky
wrote:
On Sun, Jul 29, 2012 at 3:57 PM, Josiah Carlson
wrote:

But if you're looking to perform "complicated machine learning
processing", then Redis is not the tool for you. Most machine learning
techniques rely on large matrix multiplication and/or linear
optimization, neither of which can be done efficiently with Redis.
With Redis, you are reading/writing data with a round-trip to a remote
server, which means reading/writing 100k-1M items/second (or 25k-250k
from a single client) against a single server. You won't get any
optimized algorithms for free, which means that you will not be doing
anything "complicated" with any volume of real data.

You are better off using one of the available libraries in your
language of choice, or implementing them yourself, which will let you
read/write 1B+ items/second (main memory is so much faster than a
network roundtrip), use optimized algorithms (improving the big-O
runtime), and could let you use pre-existing known-good implementation
of these algorithms.
"Out-of-core" linear algebra is expensive, which is why I brought up
the revenue issue. Pretty much the only game in town for packaged
large-scale number-crunching that doesn't involve writing a lot of
code and doesn't cost an arm and a leg is Mahout. Just about
everything else is either proprietary or falls on its hiney once it
goes beyond the capacity of a single machine's RAM and CPU.


--
Twitter: http://twitter.com/znmeb Computational Journalism Studio
http://j.mp/CompJournStudio

Data is the new coal - abundant, dirty and difficult to mine.

--
You received this message because you are subscribed to the Google Groups "Redis DB" group.
To post to this group, send email to redis-db@googlegroups.com.
To unsubscribe from this group, send email to redis-db+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/redis-db?hl=en.
--
You received this message because you are subscribed to the Google Groups "Redis DB" group.
To post to this group, send email to redis-db@googlegroups.com.
To unsubscribe from this group, send email to redis-db+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/redis-db?hl=en.

Search Discussions

Discussion Posts

Previous

Follow ups

Related Discussions

Discussion Navigation
viewthread | post
posts ‹ prev | 5 of 10 | next ›
Discussion Overview
groupredis-db @
categoriesredis
postedJul 29, '12 at 8:35p
activeJan 21, '16 at 9:08p
posts10
users6
websiteredis.io
irc#redis

People

Translate

site design / logo © 2021 Grokbase