Homework Assignments on Computability, Unsolvability, and Randomness in Math 497A, Assignments of Mathematics

The ninth homework assignment for math 497a, a university-level course on computability, unsolvability, and randomness. The assignment, authored by stephen g. Simpson, covers topics such as hoeffding's inequality, the strong law of large numbers, weakly 1-random points, and biimmune sets. Students are asked to prove various mathematical statements and theorems related to these topics.

Typology: Assignments

Pre 2010

Uploaded on 09/24/2009

koofers-user-msb
koofers-user-msb 🇺🇸

9 documents

1 / 2

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
Computability, Unsolvability, Randomness
Math 497A: Homework #9
Stephen G. Simpson
Due Monday, October 29, 2007
1. Hoeffding’s Inequality says that the probability space 2Nwith the fair
coin probability measure satisfies
Prob Pn1
i=0 X(i)
n1
2
>
!<2
exp 2n2.
Use Hoeffding’s Inequality to prove that if a point X2Nis random
(i.e., random in the sense of Martin-L¨of), then Xobeys the Strong Law
of Large Numbers:
Pn1
i=0 X(i)
n1
2as n→∞.
2. Prove that there exist weakly 1-random points in 2Nwhich do not obey
the Strong Law of Large Numbers.
Hint: Use finite approximation.
3. In problem 1, can you say anything about the rate of convergence to
1/2?
4. Prove that if XY2Nis random (i.e., random in the sense of
Martin-L¨of), then XTYand YTX.
5. Prove that there exist points X, Y 2Nsuch that XYis weakly
1-random yet XTY.
6. A set BNis said to be biimmune if both Band its complement
N\Bare immune. Prove that if X2Nis weakly 1-random then X
is the characteristic function of a biimmune set.
1
pf2

Partial preview of the text

Download Homework Assignments on Computability, Unsolvability, and Randomness in Math 497A and more Assignments Mathematics in PDF only on Docsity!

Computability, Unsolvability, Randomness

Math 497A: Homework

Stephen G. Simpson

Due Monday, October 29, 2007

  1. Hoeffding’s Inequality says that the probability space 2

N with the fair

coin probability measure satisfies

Prob

n− 1 i=

X(i)

n

exp 2n

2

Use Hoeffding’s Inequality to prove that if a point X ∈ 2

N is random

(i.e., random in the sense of Martin-L¨of), then X obeys the Strong Law

of Large Numbers:

n− 1 i=

X(i)

n

as n → ∞.

  1. Prove that there exist weakly 1-random points in 2

N which do not obey

the Strong Law of Large Numbers.

Hint: Use finite approximation.

  1. In problem 1, can you say anything about the rate of convergence to

1/2?

  1. Prove that if X ⊕ Y ∈ 2

N is random (i.e., random in the sense of

Martin-L¨of), then X T Y and Y T X.

  1. Prove that there exist points X, Y ∈ 2

N such that X ⊕ Y is weakly

1-random yet X ≡T Y.

  1. A set B ⊆ N is said to be biimmune if both B and its complement

N \ B are immune. Prove that if X ∈ 2

N is weakly 1-random then X

is the characteristic function of a biimmune set.

  1. Let f be a Turing oracle.

For each i ∈ N define

U

f i =^ {X^ ∈^2

N | ϕ

(1),f ⊕X i (0)^ ↓}.

Thus U

f i ,^ i^ = 0,^1 ,^2 ,...^ is the standard recursive enumeration of all

Σ

0 ,f 1 subsets of 2

N .

Given a sequence of sets Vn ⊆ 2

N , n = 0, 1 , 2 ,.. ., prove that the follow-

ing are pairwise equivalent.

(a) There exists a total recursive function g such that Vn = U

f g(n) for

all n.

(b) There exists a total f -recursive function h such that Vn = U

f h(n)

for all n.

(c) The predicate P ⊆ 2

N × N given by

P (X, n) ≡ X ∈ Vn

is Σ

0 ,f

In this case we say that the sequence of sets Vn, n = 0, 1 , 2 ,... is

uniformly Σ

0 ,f 1 or^ uniformly^ Σ

0 1 relative to^ f^.

Note: This concept will be part of the definition of what it means for

a point X ∈ 2

N to be random relative to the oracle f.