smasis001/strangeloop-2021

Workshop for Strangeloop Conference September 2021

Workshop for Strangeloop Conference September 2021

Mitigating Bias in Machine Learning Models with Constraints

Downloadable Pre-requisites

• Jupyter environment with *Python >= 3.6 and the following python libraries are required:

1. *pandas >= 1.0.1
2. *numpy >= 1.18.1
3. *scikit-learn >= 0.22.1
4. *scipy >= 1.5.2
5. xgboost >= 0.90
6. *matplotlib >= 3.1.3
7. *seaborn >= 0.10.0
8. tensorflow >= 2.4.1
9. pydot >= 1.3.0
10. graphviz >= 0.10.1
11. shap >= 0.35.0
12. xai == 0.0.5
13. tensorflow-lattice >= 2.0.7

(please note that Google Colab comes with all those marked with * pre-installed and code provided will come with instructions on how to install everything else)

• Dataset located here: https://github.com/PacktPublishing/Interpretable-Machine-Learning-with-Python/raw/master/datasets/recidivism-risk_balanced.csv

Additional Instructions

If you plan to run workshops code in your local machine, make sure you have a working Jupyter environment with the latest version of Python. If you don’t have one, you can install Anaconda, but please do so before the session. The code we will using is located here. Although additional libraries can be installed quickly and usually do so without any issues, it is recommended that you install them in advance (just in case the local environment presents some problems).

If you are planning to run the code in the cloud, the Colab notebook is located here. It will come with instructions on how to install the additional libraries in Colab. Please install these at the beginning of the session.