Dear Partners,

we integrated three conditional density estimators into the OpenTox framework. So now it is possible to predict intervals (an upper and a lower bound) for real-valued endpoints.
To get to know more about the conditional density estimators please have a look at Stefan's slides from Rhodes (http://opentox.org/home/documents/presentations/rhodesworkshoppressept2010/at_download/file, slides 142, 149) or at our AAAI publication from this summer: http://wwwkramer.in.tum.de/people/research/research/pubs/AAAI10_CDE.
 

With the following two curl calls you can learn and apply a model.

Learn a model:
curl -X POST -d 'dataset_uri=http://apps.ideaconsult.net:8080/ambit2/dataset/22794' -d 'prediction_feature=http://apps.ideaconsult.net:8080/ambit2/feature/140641' -d 'estimatorType=2' -d 'numBins=10' http://opentox.informatik.tu-muenchen.de:8080/OpenTox-dev/algorithm/IntervalEstimator
---> http://opentox.informatik.tu-muenchen.de:8080/OpenTox-dev/model/TUMOpenToxModel_IntervalEstimator_14

Use the learned model for predicting:
curl -X POST -d 'dataset_uri=http://apps.ideaconsult.net:8080/ambit2/dataset/22794' -d 'dataset_service=http://apps.ideaconsult.net:8080/ambit2/dataset' http://opentox.informatik.tu-muenchen.de:8080/OpenTox-dev/model/TUMOpenToxModel_IntervalEstimator_14
---> http://apps.ideaconsult.net:8080/ambit2/dataset/22844

If you have any questions please do not hesitate to contact us.


Best regards,

Fabian
-- 
GRATIS! Movie-FLAT mit über 300 Videos. 
Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome