Hi, to know what they mean you should directly go to the opencv machine learning documentation which provided information about each algorithms and related parameters(used internally by otb machine learning framework). For instance for ANN:
http://docs.opencv.org/modules/ml/doc/neural_networks.html FYI, the default value in OTB are the same as in OpenCV for most algorithms. Advises about recommended values are a far more complicated question which will depends of the learning algorithm and the data you want to classify. Le 06/02/2015 16:20, Hristo Nikolov a écrit : > Dear all, > > I try to use machine learning toolbox by OTB. I got several positive > outputs, but I need more information > on the parameters used in SVM, ANN, DT modules of the toolbox. It > would be really nice to have more > information on the parameters - what they mean and recommended values. > > Another problem I faced while I was trying to implement the learning > process of the ANN module is that I got the following error : > > > /OpenCV Error: Bad argument (output training data should be a > floating-point matrix with the number of rows equal to the number of > training samples and the number of columns equal to the size of last > (output) layer) in CvANN_MLP::prepare_to_train, file > ..\..\..\..\src\opencv-2.4.5\modules\ml\src\ann_mlp.cpp, line 702 / > /2015 Feb 06 13:21:45 : Application.logger (DEBUG) Entering DoExecute / > /2015 Feb 06 13:21:45 : Application.logger (INFO) Elevation > management: setting default height above ellipsoid to 650 meters / > /2015 Feb 06 13:21:53 : Application.logger (INFO) Number of training > samples: 3391 / > /2015 Feb 06 13:21:53 : Application.logger (INFO) Size of training > set: 3391 / > /2015 Feb 06 13:21:53 : Application.logger (INFO) Size of validation > set: 8772 / > /2015 Feb 06 13:21:53 : Application.logger (INFO) Size of labeled > training set: 3391 / > /2015 Feb 06 13:21:53 : Application.logger (INFO) Size of labeled > validation set: 8772 / > /2015 Feb 06 13:21:53 : Application.logger (FATAL) The following error > occurred during application execution : > ..\..\..\..\src\opencv-2.4.5\modules\ml\src\ann_mlp.cpp:702: error: > (-5) output training data should be a floating-point matrix with the > number of rows equal to the number of training samples and the number > of columns equal to the size of last (output) layer in function > CvANN_MLP::prepare_to_train / > > The params I used are below > > Algorithm TrainImagesClassifier (ann) starting... > C:/OSGeo4W/apps\..\bin\otbcli_TrainImagesClassifier > -io.il > "..../99_PanagReg.tif" > -io.vd > "./OTB_SVMClassification.shp" > -io.imstat > ./99_PanagReg_stats.xml > -elev.default > 650 > -sample.mt > 200 > -sample.mv > 1000 > -sample.edg > true > -sample.vtr > 0.5 > -sample.vfn > CODE_06 > -classifier > ann > -classifier.ann.t > back > -classifier.ann.sizes > 17 > -classifier.ann.f > sig > -classifier.ann.a > 1 > -classifier.ann.b > 1 > -classifier.ann.bpdw > 0.1 > -classifier.ann.bpms > 0.1 > -classifier.ann.rdw > 0.1 > -classifier.ann.rdwm > 0 > -classifier.ann.term > all > -classifier.ann.eps > 0.01 > -classifier.ann.iter > 1000 > -rand > 1 > -io.confmatout > "../ConfMatrix_2HL.txt" > -io.out > "./OutModel_2HL.txt" > > Any suggestion will be highly appreciated. > > Hr. > > -- > -- > Check the OTB FAQ at > http://www.orfeo-toolbox.org/FAQ.html > > You received this message because you are subscribed to the Google > Groups "otb-users" group. > To post to this group, send email to [email protected] > To unsubscribe from this group, send email to > [email protected] > For more options, visit this group at > http://groups.google.com/group/otb-users?hl=en > --- > You received this message because you are subscribed to the Google > Groups "otb-users" group. > To unsubscribe from this group and stop receiving emails from it, send > an email to [email protected] > <mailto:[email protected]>. > For more options, visit https://groups.google.com/d/optout. -- Manuel GRIZONNET -- -- Check the OTB FAQ at http://www.orfeo-toolbox.org/FAQ.html You received this message because you are subscribed to the Google Groups "otb-users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/otb-users?hl=en --- You received this message because you are subscribed to the Google Groups "otb-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
