The dark side of deep learning


Everyone has heard at least once of the magical powers of machine learning - who wouldn't want to be able to program a driverless car on its own? How many times, however, have you heard about the "dark sides" of data-driven technology? In this talk, we will show live demos of five aspects of deep learning that any developer should be aware of: (i) algorithmic bias; (ii) adversarial attacks on trained systems; (iii) breaches of privacy; (iv) safety threats; and (v) hidden technical debts. The lesson is: beware of blind reliance on deep learning - unless you are looking for "deep" troubles!

Language: English

Level: Intermediate

Simone Scardapane

Post-doc fellow - University of Rome "La Sapienza"

Simone Scardapane is a post-doc fellow at Sapienza University (Rome) and an honorary fellow at the University of Stirling (UK). His research is focused on machine learning, with an emphasis on deep learning, distributed environments, and applications in the audio field. Before the PhD, he obtained a B.Sc. in Computer Engineering in 2009, and a M.Sc. in Artificial Intelligence and Robotics in 2011. He is an active member of several organizations, including the IEEE Computational Intelligence Society, the International Neural Networks Society, and the AI*IA.

Go to speaker's detail