DocumentCode :
2237333
Title :
Dimension in complexity classes
Author :
Lutz, Jack H.
Author_Institution :
Dept. of Comput. Sci., Iowa State Univ., Ames, IA, USA
fYear :
2000
fDate :
2000
Firstpage :
158
Lastpage :
169
Abstract :
A theory of resource-bounded dimension is developed using gales, which are natural generalizations of martin-gales. When the resource bound Δ(a parameter of the theory) is unrestricted, the resulting dimension is precisely the classical Haludolff dimension (sometimes called “fractal dimension”). Other choices of the parameter Δ yield internal dimension theories in E, E2, ESPACE, and other complexity classes, and in the class of all decidable problems. In general, if C is such a class, then every set X of languages has a dimension in C, which is a real number dim(X|C)∈[0, 1]. Along with the elements of this theory two preliminary applications are presented: 1. For every real number 0⩽α⩽½, the set FREQ(≈α), consisting of all languages that asymptotically contain at most α of all strings, has dimension ℋ(α)-the binary entropy of α-in E and in E2. 2. For every real number 0⩽α⩽1, the set SIZE(α2n/n), consisting of all languages decidable by Boolean circuits of at most α2n/n gates, has dimension α in ESPACE
Keywords :
computational complexity; decidability; ESPACE; classical Haludolff dimension; complexity classes; decidable problems; fractal dimension; gales; internal dimension theories; natural generalizations; resource-bounded dimension; Artificial intelligence; Character generation; Computer science; Entropy; Extraterrestrial measurements; Lifting equipment; Mathematics; Power measurement;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computational Complexity, 2000. Proceedings. 15th Annual IEEE Conference on
Conference_Location :
Florence
ISSN :
1093-0159
Print_ISBN :
0-7695-0674-7
Type :
conf
DOI :
10.1109/CCC.2000.856747
Filename :
856747
Link To Document :
بازگشت