Jupyter Notebooks and supporting code for our machine learning workshop.
Given at
- TMA22 PhD School (June 2022)
- Radboud AI Study association CognAC (March 2023)
This workshops contains two assignments:
- Assignment1: Participants use this notebook to train, test and tune
scikit-learnclassifier(s) that discriminate between benign and fraud transactions. - Assignment2: Participants use this notebook to inclemently improve their classifier(s) using active learning.
We use Google Colab during the workshop. You can also run the notebooks locally by cloning this repository and setting up a Python environment yourself.
The workshop was created by Thymen Wabeke and Thijs van den Hout. Please contact us via sidnlabs@sidn.nl if you have a question.
The workshop is centered around the 'Credit Card Fraud Detection' dataset by ULB's Machine Learning Group and was originally published on Kaggle.
Some supporting code on active learning was inspired by Google's Active Learning Playground and the modAL packageA.