Talk: Uncertainty quantification with Fortuna by Gianluca Detommaso (AWS)

We are happy to welcome Gianluca Detommaso (Applied Scientist @ AWS) for an online presentation on the Fortuna library for uncertainty quantification (https://github.com/awslabs/fortuna)

The link will be posted shortly before the talk.

Speakers:

Gianluca Detommaso, Ph.D, Applied Scientist @ AWS

Timetable:

19:00 – welcome
19:05 – talk
20:00 – Q&A
20:30 – end

Title:
Uncertainty quantification with the Fortuna library

Abstract:
At AWS we released Fortuna, an open-source library for uncertainty quantification. Fortuna provides calibration methods, e.g. conformal prediction and temperature scaling, as well as several scalable Bayesian inference solutions. In this talk we will give a general overview of the field, and show how you can use Fortuna to assess the reliability of your model predictions.

Bio:

Gianluca Detommaso has been working on Bayesian inference since the beginning of his PhD in Mathematics at the University of Bath. After his PhD, he worked for two years at Amazon on probabilistic modelling for pricing applications, and over one year at AWS on uncertainty quantification (UQ) in deep learning. In his spare time, Gianluca likes to practice sports, eating great food and learning new skills.