YellowbrickTypeError for Keras Model: This estimator is not a classifier; try a regression or clustering score visualizer instead

Question

I have the following Keras DNN model and have imported necessary Keras & Yellowbrick libraries:

optimizer = RMSprop(0.001)
finalDNNModel_wrap = KerasClassifier(build_fn=parkOptimalDNN(optimizer), 
epochs=750, 
batch_size=10, verbose=0)
finalDNNModel = Sequential()
finalDNNModel.add(Dense(32, input_dim=8, activation='relu'))
finalDNNModel.add(Dense(8, activation='relu'))
finalDNNModel.add(Dense(1, activation='sigmoid'))
#Compile the model
finalDNNModel.compile(loss='binary_crossentropy',optimizer=optimizer, metrics=['accuracy'])
# Fit & Evaluate on the independent validation data set
finalDNNModel.fit(X_pca_train,y_train,batch_size = 10,epochs = 750 , verbose = 0)
dnnPrediction = (finalDNNModel.predict(X_pca_validation) > 0.5).astype("int64")
dnnPredictProb = finalDNNModel.predict_proba(X_pca_validation)

I used the YellowBrick visualisation package for classification report as below:

classes = ["Not Parkinson", "Parkinson"]
pd.set_option('precision',2)
vizDNN = ClassificationReport(finalDNNModel,classes =  classes,cmap="YlGnBu",is_fitted=True, 
force_model=True, 
title="DNN")
vizDNN.score(X_pca_validation, y_validation)
vizDNN.show()

It gives the error: "YellowbrickTypeError: This estimator is not a classifier; try a regression or clustering score visualizer instead!" Could somebody help

Answer 1

Try the following

# Import the wrap function and a Yellowbrick visualizer
from yellowbrick.contrib.wrapper import wrap
from yellowbrick.classifier import classification_report

# Instantiate the third party estimator and wrap it, optionally fitting it
finalDNNModel.fit(X_pca_train,y_train,batch_size = 10,epochs = 750 , verbose = 0)
model = wrap(finalDNNModel)

# Use the visualizer
oz = classification_report(model, X_train, y_train, X_test=X_test, y_test=y_test, support=True, is_fitted=True)

change train and test variables accordingly