Tivadar Danka

I am not sure what would be a proper solution here. Forcing bootstrapping to always contain at least two classes is kind of an artificial solution. Thinking about what to...

Hi! Thanks for the PR! I'll try to remove it soon and get back to you.

Just a quick update, I am still working on reviewing it, not sure if I understand everything correctly, so I need to spend more time with it. I have also...

To only add training data without refitting the estimator, you can use the `ActiveLearner._add_training_data` method. (Here is the implementation: [https://github.com/modAL-python/modAL/blob/master/modAL/models/base.py#L68-L92](https://github.com/modAL-python/modAL/blob/master/modAL/models/base.py#L68-L92)) This is a "private" method, so I didn't include it...

Yes, it should be deleted manually. Otherwise, the query strategy might select data which is already part of your training data, hence possibly leading to model bias in some scenarios....

Hi! I have absolutely zero knowledge regarding this question :) I don't want to state something which is false so I rather not give any advice here. Instead, I'll keep...

Hi! I am not sure I understand the issue. Are you using Keras and its scikit-learn wrapper? I suspect that your problem can be solved with writing a custom adapter...

Hi! Neural networks are supported, currently you can use Keras with its scikit-learn wrapper or PyTorch with the [Skorch](https://skorch.readthedocs.io/en/latest/index.html) library.

Hi! This is how the scikit-learn wrapper is implemented in Keras. If you check out [how its `.fit()` method is implemented](https://github.com/keras-team/keras/blob/1cf5218edb23e575a827ca4d849f1d52d21b4bb0/keras/wrappers/scikit_learn.py#L119), you can see that it calls the builder function...

Hi! The only requirement for the model is to follow the scikit-learn API. Roughly speaking, it should have `fit`, `predict` and `predict_proba` methods. If _fylearn_ learners satisfy it, in principle...