Primitive Annotation

To reduce the time needed for the tedious part of machine learning(preprocessing), we need to gain more insight in strengths and weaknesses of algorithms. What are the bottlenecks and downfalls of common machine learning algorithms?

By testing the robustness of machine learning algorithms we can rank them in efficiency. A step further is to include prepossessing steps impact on these algorithms. In this case we analyze the classification algorithms implimentation in the scikit-learn library. By learning these properties we can feed automated learn algorithms or give guidance for users to pick based on their priorities.