ML Engineering
TabArena
Individual test-row predictions from eight published TabArena methods, with derived classification correctness or regression squared error.
346,195items
384subjects
The framework is Apache-2.0. Each OpenML dataset has its own license and attribution in its captured data description and item features. These include CC BY, CC BY-SA, CC BY-NC-SA, public-domain, MIT and DbCL notices; some source descriptions state only Public. No blanket framework license applies to the data.license
generaldomain
ml_engineeringdomain
tabularmodality
item-level responses released
Response matrix
Loading response matrix…
lowhighUnobserved
Scale: Item-specific grading scales