Abstract
The efficiency of an information storage technique based on binary comparisons is analyzed. Generating functions are applied to finding the mean and variance of the number of comparisons needed to retrieve one item from a store of n items. Surprisingly, the variance approaches 7 - 2/3π2 for large n.
Original language | English (US) |
---|---|
Pages (from-to) | 376-384 |
Number of pages | 9 |
Journal | Journal of the ACM (JACM) |
Volume | 21 |
Issue number | 3 |
DOIs | |
State | Published - Jul 1 1974 |
Keywords
- binary trees
- directory search
- generating functions
- information retrieval
- mean
- searching
- trees
- variance
ASJC Scopus subject areas
- Software
- Control and Systems Engineering
- Information Systems
- Hardware and Architecture
- Artificial Intelligence