Apply min-max normalisation of SigmaNet features

Modifies the SigmaNet implementation to include optional min-max normalisation (division of values by the difference between the clamping minima and maxima) of features, required for the latest topo. trigger tuning.

To-do:

• Add min-max functionality
• Make behaviour optional, so as to not affect other SigmaNet MVAs

Cc @bldelane @nschulte

Required for Moore!4724 (merged)

Nice. LGTM.

added RTA RTA WP3 selection labels

assigned to @jagoodin

added 1 commit

• 9b3f82d4 - Make min-max normalisation optional (disabled by default)

Compare with previous version

changed the description

mentioned in merge request Moore!4724 (merged)

changed the description

marked the checklist item Add min-max functionality as completed

added 1 commit

• b9b7befe - Make min-max normalisation optional (disabled by default)

Compare with previous version

added 1 commit

• a8c29587 - Make min-max normalisation optional (disabled by default)

Compare with previous version

marked the checklist item Make behaviour optional, so as to not affect other SigmaNet MVAs as completed

requested review from @bldelane

Hi @bldelane, this should be ready now so I'd appreciate if you could take a look over this! I'll leave this in draft until Moore!4724 (merged) is ready, in case we discover we need to make any more changes in Rec.

• [2025-04-11 14:29] Validation started with lhcb-master-mr#13339
• [2025-04-16 17:06] Validation started with lhcb-master-mr#13418
• [2025-04-17 05:26] Validation started with lhcb-master-mr#13420

added 2 commits

• 9b95d8d2 - Replace | by ||
• 2f3230de - Store values for final NN inference in SigmaNet

Compare with previous version

added 1 commit

• f1467467 - Store values for final NN inference in SigmaNet

Compare with previous version

added 1 commit

• 6f57a106 - Store values for final NN inference in SigmaNet

Compare with previous version

added 1 commit

• 9325fdfb - Implement comments from Miroslav

Compare with previous version