Share it

The purpose of the workshop is to outline and exemplify the use of the AIF360 open source toolkit which has been designed to measure, understand and reduce AI bias. It combines the best bias metrics and bias mitigation algorithms.
During the workshop Its usefulness will be explained with a use case based on medical data.

Addressed to:

– Business managers preferably in healthcare – Analysts – Data science experts A good technical background can make the session easier to follow but no knowledge of programming or statistics is required.

Workshop dynamics:

The session will be led by the person running it.
Based on the information presented, there will be an activity which the participants will work on in groups of 5-8 people depending on the number of attendees.

Programa

Part I:
Description of the type of medical data. The EHR data type will be introduced with its specific features and codifications.

Part II:
Presentation of a technical study of a bias study. How are such studies approached? Who are they for?

Part III:
Group dynamics. An activity to be carried out by the participants will be presented.

Part IV:
Sharing opinions and conclusions.

Workshop given by Everis (CIDAI core partner)

Taught by:

Marilena Budan Budan, Data Scientist at NTT Data AI Center of Excellence
Mario Mesas Rodriguez, Machine Learning Engineer at NTT Data AI Center of Excellence Nohemy Pereira-Veiga Moyar, Senior Lead Data Scientist at NTT Data AI Center of Excellence
CIDAI