[R] Bigining with a Program of SVR

[ypriverol]

Well:
 I programmed an script in R using caret package and the results are very
interesting ...
 I have two datasets the first dataset have a linear distribution
experimentaly:
  values are: 4.3 , 5.3, 6.3.. 10.3... 
 the svmRadial kernel work perfectly and I can obtain an R2 = 0.98 between
the predicted value and the 
 estimated value.  
 But the second dataset do'nt have a linear distribution of the data
experimentaly ...then I can only say 
 that molecules are in the the well, 1, 2, 3, ... 10.
 When I tested the svmRadial the results are disastrous....
 Is related the problem with non-linear regression?
 Is related the problem with the fact that svmRadial in caret select the
RMSD to choice the better model?
 
Thanks in advance
Yasset  


   


--
View this message in context: http://r.789695.n4.nabble.com/Bigining-with-a-Program-of-SVR-tp3484476p3580705.html
Sent from the R help mailing list archive at Nabble.com.