Files
aimsight-libredesk/snippets/textes/how-are-targets-with-excessive-recognition-error-handled-in-data-exploration.md
T
maximeandClaude Opus 5.5 10ed887417 Noms plus clairs : snippets/, scripts de test numérotés, verifier_ollama.py
- base_connaissances/ → snippets/ (corpus/ → textes/, instructions.md → consignes_redaction.md, decoupage.py → fichiers.py, importer.py → envoyer.py) ; python3 -m snippets.envoyer.
- instance-de-test/ : 1_demarrer.sh, 2_creer_boite_et_cle.py, 3_brancher_ia.py, 4_essayer_brouillons.py, voir_empreintes.sh, connexion_admin.py.
- ollama/verifier.py → ollama/verifier_ollama.py ; tests, tâches VS Code et docs alignés.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-12 10:27:06 +01:00

16 lines
846 B
Markdown

# How are targets with excessive recognition error handled in Data Exploration?
The Data Exploration tools can use the Error Value to identify and handle measurements affected
by poor target recognition.
Available processing and visualization options include:
• Highlighting periods or areas where the Error Value is too high.
• Hiding data points whose Error Value exceeds the selected threshold.
• Replacing rejected or hidden data points with interpolated values where appropriate.
• Overriding the default Error Threshold with a user-defined value.
• Applying additional filtering or smoothing, such as a rolling median, to the resulting data
where appropriate.
The original Error Value remains useful for distinguishing actual displacement from periods
where the target was not reliably visible.
Source : AIMsight FAQ, rév. 1.5