Randomness Extraction Beyond the Classical World

Randomness is an extremely useful resource in diverse fields in computer science, such as randomized algorithms, distributed algorithms, and cryptography. In these fields, we often assume that perfect randomness are available. However, the assumption may not be realistic, since random sources available in our physical world tend to be correlated and biased. Furthermore, there is no way to ensure or verify this assumption. Randomness extraction is a natural approach to bridge the gap, where we extract (almost) perfect randomness from weak random sources and assumptions. In this talk, I will present our recent progress on several randomness extraction approaches (beyond the “classical world”).