Table 8 Results of the KM-DBSCAN data reduction compared to other methods from the related work on the collision dataset.

From: KM-DBSCAN: an enhanced density and centroid based border detection framework for data reduction towards green AI

Method

Accuracy

Loss/Gain

in Accuracy

Reduction

Ratio

Carbon (g)

Speedup

Whole dataset

0.9137

–

–

1.500

–

SRS15

0.8800

− 0.0337

0.100

0.263

9.9276

CLC15

0.8830

− 0.0307

0.100

0.556

3.8736

MMS15

0.8510

− 0.0627

0.100

0.300

8.3564

DES15

0.8920

− 0.0217

0.100

0.308

8.1657

NRMD15

0.8210

− 0.0927

0.100

0.267

9.7511

FES15

0.5830

− 0.3307

0.100

1.028

2.1457

Proposed method

0.8932

− 0.0205

0.079

0.1328

11.780