Professional Documents
Culture Documents
6. (20p) Let T = (V, E) be a tree with no vertex of degree 2 and |T | ≥ 2. Prove that T has
more leaves than other vertices.
Solution: Let n be the number of verticesPof T , and ` the number of leaves. By double
counting the edges we get that 2|E(T )| = v∈T P d(v). Since T is a tree |E| = n − 1 and
since there are no vertex of degree 2 we have v∈T d(v) ≥ 1 · ` + 3(n − `). This implies
2(n − 1) ≥ ` + 3(n − `) which is equivalent with −2 ≥ (n − `) − `. Thus there are in fact at
least 2 more leaves than other vertices.
7. (20p) Let c > 0 be a fixed constant. Does every large enough graph G, with ||G|| < c · |G|,
contain a set of 100 independent vertices?
8. Let k ≥ 2. Show that every k-connected graph of order at least 2k contains a cycle of
length at least 2k.
Solution: Let G be such a graph for which this is not the case. Let C be a cycle of
maximal length ` < 2k. There are hence at least one vertex v outside C. Let X be the
connect component that contains v in G \ C. Two neighbours in C cannot both be adjacent
to vertices in X since then there would be a longer cycle combining C with a path of length
at least 2 via X. Thus at most `/2 < k vertices in C are adjacent with some vertex in X.
But that set of vertices in C would disconnect X from the rest of the graph, contradicting
that G is k-connected.
Solution: a) Let P be the set of paths of length n − 1, which correspond bijectively to the
set of permutations of length n, thus there are n! possible Hamiltonian paths. All orienta-
tions equally likely are equivalent to the probability of each being directed each way with
probability 1/2 independent of the other edges direction. The probability that a particular
path has all the n − 1 edges directed in the direction of the path is thus (1/2)n−1 . Let X be
the random variable counting how many directed Hamiltonian paths there are and let XP be
the indicator random variable for a given path P ∈ P, i.e.
(
1, if P is a directed Hamiltonian path
XP =
0, otherwise.
By linearity of expectation we get
X X 1
E[X] = E[ XP ] = E[XP ] = n!( )n−1 .
P ∈P P ∈P
2
b) If the number of directed Hamiltonian paths is strictly less than 21−n n! for all orientations
then the expected value cannot equal to 21−n n!, which gives a contradiction. In formulas
X 21−n
X n!−1 21−n
X n!−1
E[X] = P[X = i] · i = P[X = i] · i < P[X = i]21−n n! ≤ 21−n n!,
i i=0 i=0
a contradiction.