Depends on what u r trying to do. Are u trying classification or clustering?
On Wed, Mar 4, 2015 at 1:08 AM, Raghuveer
wrote:
> Yes, you are right its was a directory. I see the part-m-0 file can
> you kindly suggest me how to run mahout on this file. Should i run
> classification or cluster
Yes, you are right its was a directory. I see the part-m-0 file can you
kindly suggest me how to run mahout on this file. Should i run classification
or clustering? Can you please share some sample.
thanks very much.
On Wednesday, March 4, 2015 11:06 AM, Andrew Musselman
wrote:
I don't have a terminal in front of me but are you sure tfidf-vectors is a
file, not a directory?
On Tuesday, March 3, 2015, Raghuveer wrote:
> I have data file of the formatsrc_ip,dest_ip,packet, bytes_transferred,
> src_port,dest_port, start_timestamp
> 71.105.62.168, 38.106.70.147, 1, 54, 557
I have data file of the formatsrc_ip,dest_ip,packet, bytes_transferred,
src_port,dest_port, start_timestamp
71.105.62.168, 38.106.70.147, 1, 54, 55704, 52747, 1341775056478
38.106.70.147, 71.105.62.168, 2, 1568, 52747, 55704, 1341775056478
Firstly the text like src_ip should be converted to a numb
I am trying to run the ItemSimilarityDriver in java (7) and getting a
NoSuchMethod ItemSimilarityDriver$.sparkConf()Lorg/apache/spark/SparkConf;
I must not have the correct version of spark. Which one works with mahout
1.0-SNAPSHOT?
Code:
ItemSimilarityDriver.main(new String[]{"-i","/transacti
Some terminology changes will help. We’re changing “cross-indicator” into
“cross-coocccurrence” indicator. Both are indicators that predict your primary
action.
Cross-cooccurrence finds the secondary interactions that are indicators or
predictors of the primary action. In your case AddToCart i
I am actually modeling charges. The main use case is when we have charges a,b
and c then we have charge d with a confidence, support, lift, etc...Then when I
get an account with some charges, I look this up against the model to see if
any charge may be missing (based on confidence)+
I was un