Hyperparameter optimization method based on dynamic Bayesian with sliding balance mechanism in neural network for cloud computing

Journal of Cloud Computing

Advances, Systems and Applications

Table 5 Comparison of different resources limits

Optimization Method	Results in different search progress (AP50 ± Variance)
Optimization Method	600 epochs	800 epochs	1000 epochs
GP (baseline)	0.468 ± 0.005	0.469 ± 0.004	0.470 ± 0.003
TPE	0.469 ± 0.004	0.473 ± 0.003	0.473 ± 0.003
BOHB	0.470 ± 0.002	0.472 ± 0.004	0.474 ± 0.002
Dynamic + Hau-PI	0.477 ± 0.002	0.479 ± 0.001	0.479 ± 0.001