Fairness in Machine Learning: are you sure there is no bias in your predictions?


Nowadays, Machine learning (ML) is used to predict nearly everything: from houses price to future criminals. Often, decisions based on ML systems have a significant impact on our life: they influence recruiters who evaluate our CV or banks that should grant us a loan. However, can we trust these systems and be sure that their predictions are fairly computed? What if a bias in the training data is amplified by these systems leading to unequal decisions? In the talk we will understand how to identify such biases and mitigate their effects.

Language: English

Level: Intermediate

Azzurra Ragone

Program Manager - Innovation and Diversity Advisor

Ingegnere Gestionale, con un Dottorato in Ingegneria dell'Informazione e la passione per la ricerca nell'ambito dell'IA, in particolar modo Machine Learning e Recommender System. Ho lavorato come ricercatrice per diversi anni, presso il Politecnico di Bari, l'Università di Trento, la University of Michigan (USA) e l'Università di Milano Bicocca. Ho lavorato presso Exprivia S.p.A come consulente di Business Intelligence per diversi aziende italiane (Generali S.p.A, Eni, Telecom). Negli ultimi due anni ha lavorato per Google nel team Developer Relations come Program Manager.

Go to speaker's detail