Hi guys, this is my first post so ill try to express me correctly.
I'm doing my degree final project (TFG) about database annalised with weka and trying to replicate it with pyweka and latter just with python with some libraries like scikit-learn ...
My problem is that one model i have to replicate use ADTree as base estimattor of a bagging problem, and i'm trying to do it in python but i don't find any information about an implementation of this ADTrees on python, just fount them on java from weka or jboost or something like that.
I tryed to do that :
---------------------------------------------------------------------------------------------------------
base_ADTree = Classifier(classname = "weka.classifier.trees.ADTree",
options = ["-B","10","-E","-3","-S","1"])
base_ADTree.build_classifier(train)
bagging_model = BaggingClassifier(base_estimator = base_ADTree,
n_estimators = 100,
n_jobs = 1,
random_state = 1)
bagging_model
-----------------------------------------------------------------------------------------------------------
So when I execute it, it returns me the model created with an ADTree done as base_estimator and when i try to evaluate the model with some scores like precision, recall, accuracy... I just obtain NaN results.
So i think obbyously it is incorrect, then is there any way to use ADTree (or other classifiers) from WEKA as estimators for Bagging or other Classifiers of scikit-learn all together?
Because i want to do a big model done on weka with scikit-learn and other python libraries.
Thank You guys.
I'm doing my degree final project (TFG) about database annalised with weka and trying to replicate it with pyweka and latter just with python with some libraries like scikit-learn ...
My problem is that one model i have to replicate use ADTree as base estimattor of a bagging problem, and i'm trying to do it in python but i don't find any information about an implementation of this ADTrees on python, just fount them on java from weka or jboost or something like that.
I tryed to do that :
---------------------------------------------------------------------------------------------------------
base_ADTree = Classifier(classname = "weka.classifier.trees.ADTree",
options = ["-B","10","-E","-3","-S","1"])
base_ADTree.build_classifier(train)
bagging_model = BaggingClassifier(base_estimator = base_ADTree,
n_estimators = 100,
n_jobs = 1,
random_state = 1)
bagging_model
-----------------------------------------------------------------------------------------------------------
So when I execute it, it returns me the model created with an ADTree done as base_estimator and when i try to evaluate the model with some scores like precision, recall, accuracy... I just obtain NaN results.
So i think obbyously it is incorrect, then is there any way to use ADTree (or other classifiers) from WEKA as estimators for Bagging or other Classifiers of scikit-learn all together?
Because i want to do a big model done on weka with scikit-learn and other python libraries.
Thank You guys.
