Talking about Science & Tech: Recent Episodes
Rolando Pablo Hong Enriquez
During the last 20 years or so I have developed a mixed scientific career between academy and industry. While busy on many projects of my current day job I have recently consider that maybe is time for me to share some of my experience, knowledge, and opinions on science & tech topics that I have found interesting or educational. Hope you enjoy it!
The Newcomb-Benford law, or the law of anomalous numbers describes some statistical observations about the distribution of the leading digits in some datasets.
This episode is an introduction to adversarial examples, a sort of "optical illusions" able to fool machine learning models. We will discuss what are adversarial examples, how can we generate them, why do they exist and we will explore if we can avoid them.