Talk in PyData Spain 2017

Talk in PyData Spain 2017

Data science for lazy people… genetics will work for you!

We love Data Science, but sometimes we have to do some manual and repetitive work before starting with the interesting and fun parts, but that will change.

TPOT is an open source tool built on top of scikit-learn for creating and optimizing machine learning pipelines. It can be considered a data science assistant. The library will automate from feature selection to parameter optimization, it is also able to do preprocessing or construct new features from existing ones.

TPOT tests a huge number of pipelines to provide you with the optimal one, this work is done with genetic algorithms. It is easy to use, has a familiar syntax if you have used Pandas or scikit-learn, and it’s very powerful. Let genetics work for you!

Slides: http://slides.com/j-diegohueltesvega/data-science-lazy-people#/
Code: https://github.com/DiegoHueltes/tpot-PyDataBCN
PyData URL: https://pydata.org/barcelona2017/schedule/presentation/36/