This hands-on workshop will introduce the main concepts of Machine Learning such as building features from raw data, fitting an estimator, evaluating predictive accuracy with cross-validation and mitigating overfitting issues. Those concepts will be illustrated by running through a typical predictive modeling pipeline involving pandas, numpy, scikit-learn and matplotlib in a Jupyter notebook.
Attendees should install the following packages:
First time users are advised to use the Anaconda distribution to get everything installed prior to attending the workshop.
Please use Python 3! Python 2 is legacy.
Basic knowledge of the Scipy ecosystem (numpy, matplotlib, jupyter...) is recommended to attend this workshop.