[R] Classification Analysis
ozric at web.de
Wed Apr 17 16:54:51 CEST 2002
is there some url/docs for more information or are this your own
thoughts - maybe for Ph.D thesis ?
Am 17.04.2002 15:22:08, schrieb "Rishabh Gupta" <rg117 at ohm.york.ac.uk>:
>Thanks For your reply
>I would really prefer to use R as I am using R for all the other exploratory analysis that I am doing on the data
In my humble opinion in R the package(Rpart) is good starting point
for this, but in fact you get no really rules. The result are more
sub-segmentation's which habit different with the classification (dependent) variable. But you can formulate rules on the basis of the results in a sophisticated way..........
>One question - i make electronic music with ComputerSoftware,too
>and i'm statistican - what are the 750 variables and what is your
>classification criterium ? (if it is not a secret !)
>I have a large of variables (750) because I have a lot of version of each
>kind of data. I analyse melody, rhythm, harmony, etc as a different stream
>of data. For each one, I calculate a different versions, intervals, contour,
>modulo, modulo of interval, etc. For each one of those I perform a different
>kind of analysis, basic statistics, complexity measurements, etc. For each
>kind of analysis there are five results each. E.g. for statistical, there is
>mean, SD, MAD, COV, etc.
>After performing all these different conversions/analysis I end up with
>about 750 variables (6 * 5 * 5 * 5)
>The classification groups are different depending on what set of MIDI files
>I am analysing. In some sets the criteria is genre, subject (who played the
>piece), composer, extract (which piece of music), etc.
100 MB und noch mehr gute Gründe! Jetzt anmelden und profitieren. Da ist mehr
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch
More information about the R-help