This directory contains notebooks that showcase FPR and FNR tradeoff which responsible to adjust the base predictions to meet the client False Positives and False Negatives requirement based on the confidence threshold.
To use the notebooks as intended, please make sure that you install the Python libraries in your virtual environment
from the ../requirements.txt
and ../requirements_notebooks.txt
files.
We recommend doing the following steps:
- Create a virtual Python environment and install all dependencies from
../requirements.txt
and../requirements_notebooks.txt
. - Add the virtual environment's kernel to iPython's kernel list - see these instructions for how to do this.
- Start Jupyter notebook server/Jupyter lab and use the added kernel that will include the installed dependencies.