Skip to main content

Table 9 The experimental comparison results of running time with Spark

From: Big data decision tree for continuous-valued attributes based on unbalanced cut points

Methods

Datasets

Gaussian1

Gaussian2

SUSY

HEPMASS

Covertype

BA-CDT-SP

15.4

31.2

228.2

506.5

568.5

BS-CDT-SP

19.8

35.3

312.5

572.2

626.6

Parallel C4.5-SP

31.5

45.1

325.1

687.6

597.1

MLlib-DT-SP

55.2

59.8

414.0

709.5

583.4

FRBDT

34.1

40.5

219.8

556.7

572.1

IS-C4.5

35.5

43.1

252.0

619.5

595.9

  1. The data in bold indicate the shortest run times for different algorithms implemented by MapReduce and Spark on different datasets, respectively