Table 6 Learned Bloom filters average reject time, expressed as ratio between learned filter and baseline BF reject times (whose time in seconds is in parentheses)
From: The role of classifiers and data complexity in learned Bloom filters: insights and recommendations
Classifier | LBF | SLBF | ADA-BF |
|---|---|---|---|
Synthetic Data (\(1.364\cdot 10^{-5}\)) | |||
SVM | 18.4 | 6.1 | 151.2 |
DT | \(-\)1.1 | \(-\) 1.2 | 1.3 |
RF | \(-\)11.1 | \(-\) 17.5 | 112.8 |
NN | 106.9 | 54.1 | 159.3 |
URL Data (\(3.259\cdot 10^{-5}\)) | |||
SVM | 22.6 | 3.7 | 3.9 |
DT | \(-\) 1.4 | \(-\) 1.4 | \(-\)1.1 |
RF | 6.6 | 7.1 | 9.7 |
NN | 43.9 | 49.6 | 35.3 |
DNA Data (\(4.817\cdot 10^{-5}\)) | |||
SVM | \(-\) 12.5 | \(-\)11.7 | 35.9 |
DT | \(-\)1.1 | \(-\) 1.3 | 1.3 |
RF-10 | 1.4 | \(-\) 20.6 | 32.0 |
RF-100 | 19.8 | \(-\) 7.4 | 40.6 |
NN-7 | \(-\)5.0 | \(-\) 12.0 | 25.8 |
NN-125,50 | \(-\)3.6 | \(-\) 7.5 | 25.2 |
NN-500,150 | \(-\)1.9 | \(-\) 11.6 | 39.2 |