Table 4 Speedup and memory usage comparison between ti\(k\)-NN (B+, \(ball\)-tree) and iDistance.

From: Telescope indexing for k-nearest neighbor search algorithms over high dimensional data & large data sets

Speedup achieved and Memory usage by ti\(k\)-NN (B+ + \(ball\)-tree) against iDistance

  

census

covtype

poker

MNIST

swarm

gisette

(B+ + \(ball\)-tree)

Speedup

6.27

16.78

15.23

3.09

3.79

3.16

Memory

0.92

0.91

0.90

0.93

0.96

0.92

Back to article page