Speaker: Brandon Fu
University of Notre Dame
Will give a Graduate Student Geometry Seminar entitled:
A Concentration Inequality
Abstract: In probability theory, the law of large number states that the empirical average of a large number of independent and identical random samples converges to the true underlying mean, almost surely. The question is how much this empirical average deviates from the true mean. Concentration inequality provides a solution. In this talk, we will go through Hoeffding’s inequality based on the Chernoff method and discuss Poincare phenomenon, i.e. Under some conditions, certain random vectors are highly concentrated around the unit sphere in high dimensional space. Finally, we will discuss its applications in machine learning.
Date: 04-19-2024
Time: 2:00 pm
Location: 117 Hayes-Healy Bldg