| Type of Document |
Dissertation |
| Author |
Abu-Mostafa, Yaser Said
|
| URN |
etd-05142003-111719 |
| Persistent URL |
http://resolver.caltech.edu/CaltechETD:etd-05142003-111719 |
| Title |
Complexity of information extraction |
| Degree |
PhD |
| Option |
Electrical Engineering |
| Advisory Committee |
| Advisor Name |
Title |
| A. Kechris |
Committee Member |
| C. Mead |
Committee Member |
| D. Psaltis |
Committee Member |
| E. Posner |
Committee Member |
| H. Ryser |
Committee Member |
| R. McEliece |
Committee Member |
| R. Wilson |
Committee Member |
|
| Keywords |
|
| Date of Defense |
1983-05-23 |
| Availability |
unrestricted |
Abstract
This thesis describes a mathematical theory that interrelates the basic concepts of complexity, cost, information and reliability. The accessibility of information, as opposed to its availability, is characterized. Universal bounds for complexity distribution, implementation cost and decision reliability are estimated. These bounds give rise to a methodology for any consistent definition of a complexity measure. The basic notions of pattern recognition and information theory are directly related to computational complexity.
|
| Files |
| Filename |
Size |
Approximate Download Time
(Hours:Minutes:Seconds)
|
| 28.8 Modem |
56K Modem |
ISDN (64 Kb) |
ISDN (128 Kb) |
Higher-speed Access |
| |
Abu-Mostafa_ys_1983.pdf |
2.55 Mb |
00:11:47 |
00:06:03 |
00:05:18 |
00:02:39 |
00:00:13 |
|
