> I'm sorry I didn't put axes on the graph... I think you have been misreading
> it, as you are characterizing the fluctuations as "noise."  

You are correct, I assumed that the X axis represented document size, so 
my interpretation of your graph was totally incorrect.

> Let me clarify:  The vertical axis is actual or predicted latency.
> The horizontal is just "sample index".

Why are the first few BDA1 predictions along the X axis all the same in
the first graph?

> To clarify this misunderstanding further, I've introduced another graph
> on the webpage showing file size vs. latency in inp.dat.

That certainly makes more sense.

> > I really feel that the only fair comparison will be with a large
> > realistic dataset, and being compared against the actual algorithm we 
> > have developed for this purpose, for which even BDA3 is a primitive 
> > substitute.  I will try to extract such a dataset from Hui so that we 
> > can make some progress here.
> Ok this will be the best test.

Agreed - I suggest we suspend our debate of this sample data until we 
have the results of a test with SVM and RTE using actual collected data.  
If I don't hear back from Hui soon I will see if I can collect some more 
sample data which I can use.

Ian.

-- 
Ian Clarke                                                  [EMAIL PROTECTED]
Coordinator, The Freenet Project              http://freenetproject.org/
Founder, Locutus                                        http://locut.us/
Personal Homepage                                   http://locut.us/ian/

Attachment: pgp00000.pgp
Description: PGP signature

Reply via email to