Additional classification and regression algorithms based on the scikit-learn library.

Currently implemented classes are:

• StreamingDecisionTreeClassifier
• StreamingRandomForestClassifier
• StreamingExtraTreesClassifier

Their use and parameters are essentially identical to their non-streaming originals, except that these implement partial_fit().

Note, that for datasets that can be trained in-memory with a single batch, these will perform no better than the originals. Their only benefit lies when training on a dataset too large to fit into memory, requiring the dataset to be split into smaller parts.