Python Code & Libraries


Scikit-Learn is easily the most important machine learning library in Python. It contains almost every possible algorithm, and its interface is being reused by many libraries out there.


Tensorflow is Google’s deep learning framework. It is very popular and quite easy to learn. It offers more low-level changes and is great for deployment.


SpaCy is a wonderful natural language processing package which offers great performance. Its community is constantly growing and the core developers offer great support.


Pandas offers a lot of great data manipulation tools. It is a great package when it comes to reading and writing data to files in plenty of formats.

ML From Scratch

Erik Linder-Norén wrote a lot of machine learning algorithms in Python from scratch as well.


Keras is a famous deep learning library. It is a wrapper on top of Tensorflow, which allows one to get started with deep learning with only a few lines of code.


Pytorch is Facebook’s deep learning framework and Tensorflow’s biggest rival. It has the advantage to be fully written in Python and allows for dynamic computational graphs.


NLTK is a deeply comprehensive natural language processing library. As opposed to SpaCy, it keeps a high focus on modularity and offers users a lot of freedom.


NumPy is a scientific package that allows you to perform plenty of array and vector operations. It uses a lot of lower level code which makes it quite fast.