"LibSVM classfication: X-validation evaluation , Normalization"
aryan_hosseinza
MemberPosts:74
Contributor II
Hi everybody ,
I have a dataset with 12 features for its instances (11 regular features : integers, 1 label feature : nominal)
It's a very imbalanced dataset, I want to use libSVM as a one-class classifier,
Firstly,
As you can see in the code below , I want to use X-Validation to train one-class classifier on minor-class instances of training portion(90%) of dataset and I want to apply this model on the rest of the data (10%) in each validation,
but I can't use Binomial Classification Performance evaluator , as the label is nominal ! if I want to change it to binomial I won't be able to train SVM classifier ! what should I do ?
Secondly,
Generally , what preprocessing do you think is necessary to do before using a libSVM classifier ? I read somewhere in the forum that it's necessary to do a normalization :
"Is there something like a "NormalizationModel" to get the same normalized values for trainig and test examples?
This should be possible by applying the operator Normalization with enabling the parameter return_preprocessing_model on the training data. This means, a preprocessing model is generated by normalizing the training which you can subsequently apply on the test data as well. The normalization is then done on the test data via the same transformation as applied on the training data. Therefore you have to apply this model using the ModelApplier."
but I didn't get what to do ? I would be thankful if you could help me,
And finally ,
I heard that one-class classifiers don't work very good comparing to other classifiers , is that true?
Thanks ,
Arian
I have a dataset with 12 features for its instances (11 regular features : integers, 1 label feature : nominal)
It's a very imbalanced dataset, I want to use libSVM as a one-class classifier,
Firstly,
As you can see in the code below , I want to use X-Validation to train one-class classifier on minor-class instances of training portion(90%) of dataset and I want to apply this model on the rest of the data (10%) in each validation,
but I can't use Binomial Classification Performance evaluator , as the label is nominal ! if I want to change it to binomial I won't be able to train SVM classifier ! what should I do ?
Secondly,
Generally , what preprocessing do you think is necessary to do before using a libSVM classifier ? I read somewhere in the forum that it's necessary to do a normalization :
"Is there something like a "NormalizationModel" to get the same normalized values for trainig and test examples?
This should be possible by applying the operator Normalization with enabling the parameter return_preprocessing_model on the training data. This means, a preprocessing model is generated by normalizing the training which you can subsequently apply on the test data as well. The normalization is then done on the test data via the same transformation as applied on the training data. Therefore you have to apply this model using the ModelApplier."
but I didn't get what to do ? I would be thankful if you could help me,
And finally ,
I heard that one-class classifiers don't work very good comparing to other classifiers , is that true?
Thanks ,
Arian
< portSpacing port="source_training" spacing="0"/>
< portSpacing port="sink_model" spacing="0"/>
< portSpacing端口= " sink_through 1”间隔= " 0 " / >
< portSpacing port="source_model" spacing="0"/>
< portSpacing port="source_test set" spacing="0"/>
< portSpacing port="source_through 1" spacing="0"/>
< portSpacing port="sink_averagable 1" spacing="0"/>
< portSpacing port="sink_averagable 2" spacing="0"/>
< portSpacing port="source_input 1" spacing="0"/>
< portSpacing port="sink_result 1" spacing="0"/>
< portSpacing port="sink_result 2" spacing="0"/>
Tagged:
0
