Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2506.04026 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916778721935360 |
|---|---|
| author | Bénesse, Clément Mesana, Patrick Gautier, Athénaïs Gambs, Sébastien |
| author_facet | Bénesse, Clément Mesana, Patrick Gautier, Athénaïs Gambs, Sébastien |
| contents | In machine learning, knowing the impact of a given datum on model training is a fundamental task referred to as Data Valuation. Building on previous works from the literature, we have designed a novel canonical decomposition allowing practitioners to analyze any data valuation method as the combination of two parts: a utility function that captures characteristics from a given model and an aggregation procedure that merges such information. We also propose to use Gaussian Processes as a means to easily access the utility function on ``sub-models'', which are models trained on a subset of the training set. The strength of our approach stems from both its theoretical grounding in Bayesian theory, and its practical reach, by enabling fast estimation of valuations thanks to efficient update formulae. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_04026 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | On the Usage of Gaussian Process for Efficient Data Valuation Bénesse, Clément Mesana, Patrick Gautier, Athénaïs Gambs, Sébastien Machine Learning In machine learning, knowing the impact of a given datum on model training is a fundamental task referred to as Data Valuation. Building on previous works from the literature, we have designed a novel canonical decomposition allowing practitioners to analyze any data valuation method as the combination of two parts: a utility function that captures characteristics from a given model and an aggregation procedure that merges such information. We also propose to use Gaussian Processes as a means to easily access the utility function on ``sub-models'', which are models trained on a subset of the training set. The strength of our approach stems from both its theoretical grounding in Bayesian theory, and its practical reach, by enabling fast estimation of valuations thanks to efficient update formulae. |
| title | On the Usage of Gaussian Process for Efficient Data Valuation |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2506.04026 |