[rrd-users] Data Mining: Correlation Engine
fcocquyt at stanford.edu
Wed Nov 12 04:13:24 CET 2008
Well, we are always trying to "find the signal in the noise" and it seems
this would be a natural progression of this concept.
So can any of the statistical majors in the group recommend a wickedly fast
correlation source code collection?
Since we are talking about an O(n^2) dataset time will be of the essence -
I've already seen some FFT implementations - I am leaning to compiled C code
for speed but I'd appreciate any pointers...the data is waiting to be mined
William R. Lorenz wrote:
> Hey Fletcher,
> It seems you have some very interesting ideas!
> On Mon, 10 Nov 2008, fcocquyt wrote:
>> I searched the forums without an answer - has anyone looked at a data
>> mining engine for RRDTool data? An example application would be
>> computing the correlation of different datasources in the set of all
>> datasources [...] thinking much of the untapped value of RRDtool
>> datasets rests in the analysis across the rrd files (eg wow, our online
>> transactions (sales) drop off dramatically with our backend DB latency -
>> Anyone else see value in such a data mining engine for RRDTool?
> It seems such an endeavor would involve quite a bit of inferential (as
> opposed to descriptive) statistics. As I see you have a stanford.edu
> email address, I gather you're on top of such things. :-) How would you
> propose such multi-variate time series data could be best correlated and
> statistically analyzed, across many user-friendly rrdtool deployments?
> Very interesting -- I'd like to hear more about your ideas!
> William R. Lorenz
> rrd-users mailing list
> rrd-users at lists.oetiker.ch
View this message in context: http://n2.nabble.com/Data-Mining%3A-Correlation-Engine-tp1483820p1488120.html
Sent from the RRDtool Users Mailinglist mailing list archive at Nabble.com.
More information about the rrd-users