What is scikit-learn?

Scikit-learn has been the lingua-franca of machine learning community for over a decade now. Developed in early 2000’s, the paper about scikit-learn came in 2011 and 2013. The paper is well written, and describes the basic principles of the library. The very fact that the paper of a machine learning library is still very much relevant today, speaks volumes about the strong principles and robust software design of scikit-learn library. No doubt, the library has influenced many onward machine learning libraries and almost all the ‘mainstream’ machine learning libraries have borrowed many concepts from it. Since the success of scikit-learn, there has been plethora of scikits, sickit-optimize for optimization of hyperparameters, scikit-image for image processing, to name a few. This post has been influenced by the original paper and my experience of using scikit-learn. The purpose is to provide an overview of scikit-learn with code examples.

estimators

predictors

transformers

model selection

pipelines

limitations

extending