Error loading Model

[iAndrea]

When i have to create a model for the first time there are no problem. After the creation, i use the tool with the --model option (also with malware and benign apk parameters), but after: SVMModels = load(Model) Gives me an error in: BestModel= SVMModels.best_estimator

Ok, i modify it as: BestModel= SVMModels.best_estimator_ and it works great, but the next predict doesn't work: y_pred = SVMModels.predict(x_test)

Should i retrain the model after loading, before predict? If yes, why you didn't do it...

[zwj2017-NK]

Although the program includes model file and can load it, can't work in the process of predict. You need save "FeatureVectorizer" variable during training your model. When you want to predict test data, loading the matrix variable into memory is necessary for the next stage, then,using your model and the matrix variable can be effective.

[MadDroid-2020]

Although the program includes model file and can load it, can't work in the process of predict. You need save "FeatureVectorizer" variable during training your model. When you want to predict test data, loading the matrix variable into memory is necessary for the next stage, then,using your model and the matrix variable can be effective.

Hi, can you elaborate on how to use the saved model and how to save and load "FeatureVectorizer" variable on new test apks? Many thanks in advance.