Table 2 Experiment timing comparison: personalized vs. baseline federated learning (in min)

From: Personalized federated learning for predicting disability progression in multiple sclerosis using real-world routine clinical data

Method

Experiment Time (min) ↓

 

Adaptive

Baseline

FedAVG

206.81 ± 3.670

192.66 ± 5.370

FedProx

217.51 ± 13.18

171.77 ± 9.440

FedAdagrad

230.91 ± 18.27

190.02 ± 2.240

FedYogi

246.74 ± 5.420

193.21 ± 41.89

FedAdam

225.69 ± 7.910

236.08 ± 13.42

  1. Where lower timing values are preferred, bold is used to highlight the best timing for each model type (i.e., the better-performing federated strategy between Adaptive and Baseline for that specific method). In cases where two values are very close and not meaningfully different, both are bolded to reflect comparable performance.
Back to article page