Instructor: László Babai
Scribes: Justin Sinz and Travis
Schedler
Note: (C) and (B) imply (A).
For a set of
positive real numbers,
we can define
several means:
;
;
.
Note: A continuous function satisfying inequality (
)
is concave and a continuous function satisfying
inequality (
)
is convex. However, not every function satisfying either (or even
both)
inequalities is continuous. (Prove!
)
We proved in class the
In fact we showed that,
as
tends to infinity,
.
The following exercise strengthens this result:
The following exercise shows the tightness of the bound:
Erdos proved this in 1960 using the probabilistic method but did not exhibit any specific examples of such graphs. No explicit construction was known of such graphs until 30 years later, around 1990.