Skip to main content

Advances, Systems and Applications

Table 7 Comparative Assessment of Deep-ConvLSTM_FPO Scheme With Task Size 200

From: Task grouping and optimized deep learning based VM sizing for hosting containers as a service

Number of Iteration

VM size selection technique

Hurst exponent+Markov transition

ICLB

OM-FNN

VM size IaaS multi-tenant public cloud

Deep-ConvLSTM GD

Deep-ConvLSTM ADAM

Proposed Deep-ConvLSTM_FPO

Resource Utilization

5

0.0733

0.0726

0.0725

0.0708

0.0683

0.0674

0.0669

0.0654

10

0.0792

0.0777

0.0764

0.0760

0.0758

0.0756

0.0749

0.0733

15

0.0944

0.0794

0.0779

0.0764

0.0757

0.0747

0.0737

0.0714

20

0.1040

0.0916

0.0893

0.0826

0.0793

0.0765

0.0758

0.0721

Response Time

5

34856

28852

23345

15678

11846

921

621

540

10

36799

30796

25355

15710

12805

956

658

449

15

35340

31337

23719

15846

11891

999

692

442

20

36451

32448

24747

15937

11815

1025

734

427

Task Rejection Rate

5

0.2999

0.2945

0.2914

0.2901

0.2895

0.2755

0.2698

0.2054

10

0.3587

0.3413

0.3321

0.3157

0.3026

0.2968

0.2877

0.2241

15

0.3699

0.3585

0.3486

0.3268

0.3143

0.3077

0.2976

0.2385

20

0.3413

0.3399

0.3325

0.3326

0.3326

0.3269

0.3179

0.2486

Makespan

5

0.6867

0.6467

0.6076

0.5976

0.5865

0.5778

0.5725

0.5656

10

0.7076

0.6578

0.6145

0.6096

0.6076

0.5967

0.5935

0.5865

15

0.7156

0.6765

0.6567

0.6467

0.6356

0.6256

0.6156

0.6076

20

0.7267

0.6976

0.6677

0.6576

0.6478

0.6436

0.6400

0.6356