This course explored the exciting field of homomorphic encryption and its applications to ML. Taught jointly with an expert in Mexican digital invoicing, it was a two-hour course at the Mexican Research and Applications in AI summer school.
The Jupyter notebooks for the course are available in GitHub and they run directly in Google Collab.
The code builds a blind recommendation system , that is, an ML that recommends products without knowing what products the customers have purchased.
The workshop itself (in Spanish) was recorded and it is available in YouTube. A shorter version without the hands-on bit was also presented as an on-line seminar at FaMAF and it is also available on YouTube. The slides are also available and they are in English.