
Lecture 20: Independence
Erik Demaine teaches probability theory in this MIT lecture from 6.1200J Mathematics for Computer Science, covering three forms of independence that recur throughout discrete math and computer science: mutual independence, pairwise independence, and conditional independence. The lecture works through definitions and examples showing how events can be pairwise independent without being mutually independent, a distinction that trips up many students, and how conditioning on new information changes independence relationships. As with the rest of the course, the session builds on earlier material on probability spaces and random variables, developing the formal tools needed for later topics like expectation and concentration bounds. Demaine works through the material on a whiteboard at a measured pace typical of MIT's OpenCourseWare lectures, favoring worked examples over abstract statement alone. Runtime runs 82 minutes, consistent with a full lecture session rather than a clipped excerpt.