Groups
Category
Concentration inequalities give high-probability bounds that random outcomes stay close to their expectations, even without knowing the full distribution.
Randomized algorithms use random bits to make choices that simplify design, avoid worst cases, and often speed up computation.