Effective Strong Dimension in Algorithmic Information and Computational Complexity
read more
Citations
Dimensions of Points in Self-Similar Fractals
Effective fractal dimensions
Entropy rates and finite-state dimension
Constructive dimension equals Kolmogorov complexity
Effective fractal dimension: foundations and applications
References
A mathematical theory of communication
Probability and Measure
Fractal Geometry: Mathematical Foundations and Applications
An Introduction to Kolmogorov Complexity and Its Applications
Compression of individual sequences via variable-rate coding
Related Papers (5)
Frequently Asked Questions (16)
Q2. What is the entropy rate of a type investigated by Staiger?
Constructive strong dimension can also be used to characterize entropy rates of the type investigated by Staiger [37, 38] and Hitchcock [16].
Q3. What is the key to all these effective dimensions?
The key to all these effective dimensions is a simple characterization of classical Hausdorff dimension in terms of gales, which are betting strategies that generalize martingales.
Q4. What is the binary entropy function in this paper?
N | tj ≥ n}, where t0 = 0 and tj+1 = 2tj , and Shannon’s binary entropy function H : [0, 1] → [0, 1] defined byH(β) = β log 1 β + (1− β) log 1 1− β ,where 0 log 10 = 0.
Q5. What is the second part of the proof of Lemma 7.2?
Their proof of Theorem 7.2 also uses the martingale dilation technique, which was introduced by Ambos-Spies, Terwijn, and Zheng [2] and extended by Breutzmann and Lutz [4].
Q6. What is the standard enumeration of a language?
The authors usually identify a language A with its characteristic sequence χA ∈ C defined by χA[n] = if sn ∈ A then 1 else 0, where s0 = λ, s1 = 0, s2 = 1, s3 = 00, . . . is the standard enumeration of {0, 1}∗.
Q7. what is the f -dilation of d?
If f : {0, 1}∗ −→ {0, 1}∗ is strictly increasing and ~β is a bias sequence, then the f -dilation of ~β is the bias sequence ~βf given by βfn = βnf for all n ∈ N.Observation 7.5.
Q8. What is the proof of Lemma 7.4?
Since s < H(α) = H−(~γ) = H−(~γ′),t < H(β) = H+(~γ) = H+(~γ′),and ~γ′ is O(n)-time-computable, Lemma 7.4 tells us that B 6∈ S∞[d−] and B 6∈ S∞str[d+].
Q9. What is the proof of the k-state information-lossless finite-state?
For each k ∈ N let C′k be the set of all k-state information-lossless finite-state compressors whose output length per input bit is bounded by k.
Q10. What is the stability of finite-state dimensions?
The finite-state dimensions are stable, i.e., for all X, Y ⊆ C,dimFS(X ∪ Y ) = max{dimFS(X),dimFS(Y )}and DimFS(X ∪ Y ) = max{DimFS(X),DimFS(Y )}.Proof.
Q11. What is the proof of theorem 7.6?
It follows by Theorem 7.6 that gk d̂ is a ~β-martingale that succeeds on R. The time required to compute gk d̂(w) is O(|w|2 + |w′|k) steps, where w′ = w range(gk).
Q12. What is the c-regular expansion of the real number sA?
For each s ∈ (0, 1] and each infinite, computably enumerable set A ⊆ {0, 1}∗, the (binary expansion of the) real number θsA is c-regular with dim(θsA) = Dim(θsA) = s.
Q13. What is the log-loss rate of on w isLlog(?
The cumulative log-loss of π on w isLlog(π,w) = |w|−1∑ i=0 log 1 π(w[0..i− 1], w[i]) .2. The log-loss rate of π on S isLlog(π, S) = lim inf n→∞ Llog(π, S[0..n− 1]) n .3.
Q14. What is the sequence S that has wn v S?
Their sequence S is the unique one that has wn v S for all n.Dai, Lathrop, Lutz, and Mayordomo [7] investigated the finite-state compression ratio ρFS(S), defined for each sequence S ∈ C to be the infimum, taken over all information-lossless finite-state compressors C (a model defined in Shannon’s 1948 paper [36]) of the (lower) compression ratioρC(S) = lim inf n→∞ |C(S[0..n− 1])| n .
Q15. what is the d-Lyapunov exponent of s?
The lower d-Lyapunov exponent of S is λd(S) = lim infn→∞ log d(S[0..n−1])n .2. The upper d-Lyapunov exponent of S is Λd(S) = lim supn→∞ log d(S[0..n−1]) n .3.
Q16. What is the key to the spectrum of effective dimensions?
Given this characterization, it is a simple matter to impose computability and complexity constraints on the gales to produce the above-mentioned spectrum of effective dimensions.