Bureaucrats, dave_user, Administrators
4,650
edits
Changes
→Final choice
summarised in Table 3.1:
{|
|+ FL frameworks table comparison
!align="center"| '''Framework'''!align="center"| '''NVFlare'''!align="center"| '''FATE'''!align="center"| '''Flower'''!align="center"| '''PySyft'''!align="center"| '''IBMFL'''!align="center"| '''OpenFL'''!align="center"| '''TFF'''!align="center"| '''FedML'''
|-
|align="center"| '''ML frameworks flessibility'''|align="center"| high|align="center"| fair|align="center"| high|align="center"| fair|align="center"| high|align="center"| high|align="center"| low|align="center"| high
|-
|align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"|
|-
|align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"|
|-
|align="center"| '''License'''|align="center"| Apache|align="center"| Apache|align="center"| Apache|align="center"| Apache|align="center"| Apache|align="center"| Apache|align="center"| Apache|align="center"| Apache
|-
|align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"|
|-
|align="center"| '''Repo stars'''|align="center"| 413|align="center"| 5.1k|align="center"| 2.8k|align="center"| 8.9k|align="center"| 438|align="center"| 567|align="center"| 2.1k|align="center"| 3.1k
|-
|align="center"| '''Releases'''|align="center"| Nov 23|align="center"| Feb 18|align="center"| Nov 11|align="center"| Jan 19|align="center"| Aug 28|align="center"| Feb 1|align="center"| Feb 20|align="center"| Apr 30
|-
|align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"|
|-
|align="center"| '''Documenta- tion and tutorials'''|align="center"| good|align="center"| decent|align="center"| good|align="center"| bad|align="center"| good|align="center"| bad|align="center"| good|align="center"| bad
|-
|align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"|
|-
|align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"|
|-
|align="center"| '''Readiness for commer- cial usage'''|align="center"| ready|align="center"| ready|align="center"| ready|align="center"| not ready|align="center"| not ready|align="center"| ready|align="center"| ready|align="center"| not ready
|-
|align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"|
|-
|align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"||align="center"|
|}
{| class="wikitable"|+!!!!|-|||||-|||||-|||||}
These two remaining frameworks are then: Flower and NVFlare. They demonstrated the potential to address the research objectives effectively and were