Ethics and Deep Learning
Chiara Criscuolo
DEIB, PHD Student in Information Technology
Research assistant
DEIB PT1 Room (Building 20)
October 30th, 2023
10.00 am
Contacts:
Chiara Criscuolo
Research Line:
Data, web, and society
DEIB, PHD Student in Information Technology
Research assistant
DEIB PT1 Room (Building 20)
October 30th, 2023
10.00 am
Contacts:
Chiara Criscuolo
Research Line:
Data, web, and society
Sommario
On October 30th, 2023 at 10.00 am Chiara Criscuolo will held a seminar on "Ethics and Deep Learning" in DEIB PT1 Room (Building 20).
Decision-making algorithms and systems based on big data have become essential tools that pervade all aspects of our daily lives; in order for these technologies to be reliable, the results should not only be accurate, but also ethical. Nowadays, we are not only talking about machine learning techniques, but also about the more obscure deep learning, so it is necessary to study the risks involved in using these new technologies, and how to make more more ethical. In fact, due to data bias, errors, and other factors, these technologies may provide a distorted view of reality, leading to wrong analyses, results and, consequently, decisions. This talk presents various methods to study ethical issues in deep learning, specifically to discover and mitigate fairness.
Decision-making algorithms and systems based on big data have become essential tools that pervade all aspects of our daily lives; in order for these technologies to be reliable, the results should not only be accurate, but also ethical. Nowadays, we are not only talking about machine learning techniques, but also about the more obscure deep learning, so it is necessary to study the risks involved in using these new technologies, and how to make more more ethical. In fact, due to data bias, errors, and other factors, these technologies may provide a distorted view of reality, leading to wrong analyses, results and, consequently, decisions. This talk presents various methods to study ethical issues in deep learning, specifically to discover and mitigate fairness.