insufficient results with M5P

Hello,

if anyone is interested please try the following:

produce a file containing two columns

 x = 0, 0.1, 0.2, ..., 12.6; 
 y = sin(x)

Then apply M5P.

The result is quite disappointing. Does anyone know ow to get an acceptable result?
I expected to get something like a picewise linear approximation of the sin function,
but got something far away from this.

Applying the M5P algorithm to this dataset with the default parameters does yield a piecewise linear approximation of the sin function. If you use the LinePlot (Category:Views) you can see that very well.
I obtained the best results with the following settings:
buildRegressionTree: True
minNumInstances:4.0
unpruned:False
useUnsmoothed:False

M5P is a weka node, if you have any questions about specific details of this algorithm, please refer to the weka mailing list and the following 2 papers:
Ross J. Quinlan: Learning with Continuous Classes. In: 5th Australian Joint Conference on Artificial Intelligence, Singapore, 343-348, 1992.

Y. Wang, I. H. Witten: Induction of model trees for predicting continuous classes. In: Poster papers of the 9th European Conference on Machine Learning, 1997.

  • Nicolas