Skip to main content

Advances, Systems and Applications

Table 10 Prediction performance of Bitbrains dataset for the proposed models compared to the models in [44]

From: VTGAN: hybrid generative adversarial networks for cloud workload prediction

Method

Window size

Train:Test ratio

MAPE

Bi-LSTM [44]

 

65:35

12.0119

  

70:30

12.2173

 

30

75:25

12.3019

  

80:20

13.6177

  

65:35

11.7046

  

70:30

11.7091

 

60

75:25

11.914

  

80:20

13.6163

  

65:35

12.0244

  

70:30

12.3091

 

90

75:25

12.8671

  

80:20

13.1198

  

65:35

12.0802

  

70:30

11.8903

 

120

75:25

14.207

  

80:20

13.4428

BHyPreC [44]

 

65:35

11.1799

  

70:30

12.3343

 

30

75:25

12.3688

  

80:20

12.2959

  

65:35

11.1101

  

70:30

13.0751

 

60

75:25

11.7641

  

80:20

13.507

  

65:35

12.537

  

70:30

12.2912

 

90

75:25

10.8557

  

80:20

12.4713

  

65:35

12.2044

  

70:30

10.7738

 

120

75:25

12.706

  

80:20

13.3193

VTGAN (LSTM-based)

 

65:35

10.5822

  

70:30

9.47898

 

30

75:25

9.39637

  

80:20

9.0233

  

65:35

10.911

  

70:30

10.1507

 

60

75:25

10.4705

  

80:20

9.3998

  

65:35

10.3466

  

70:30

13.6877

 

90

75:25

11.146

  

80:20

11.2193

  

65:35

13.0493

  

70:30

14.7279

 

120

75:25

12.4581

  

80:20

12.8819

VTGAN (GRU-based)

 

65:35

8.87

  

70:30

8.6018

 

30

75:25

9.1799

  

80:20

9.6228

  

65:35

8.5347

  

70:30

8.4522

 

60

75:25

9.044

  

80:20

8.1686

  

65:35

8.747

  

70:30

8.8152

 

90

75:25

8.5942

  

80:20

8.3724

  

65:35

8.6346

  

70:30

9.0875

 

120

75:25

8.0545

  

80:20

8.5751