Sample-data demonstration
This shows our AI testing process on public sample data. You can't upload your own data here.
From YourCloudGroup Model Manager.

Question: Who will say yes to an offer
Sample data: Bank Marketing (UCI)
Attribution: Moro, S., Rita, P., & Cortez, P. (2014). Bank Marketing [Dataset]. UCI Machine Learning Repository. https://doi.org/10.24432/C5K306. CC BY 4.0. Modified: split into training and held-out rows, predictions added.
Predicts: y (subscribed) (Classification)

Result
Its predictive value is WEAK, measured but not clearly better than the baseline.
Certification: CERTIFIED — AUDITABLE & REPEATABLE PREDICTIVE MODEL
Value verified: no. The winning model showed limited improvement over the baseline; production value is not established.
Winner: XGBoost, best of 12 models built
Trained on: 10,000 rows (the same 10,000-row sample of the 40,690 training rows for every model)

Certificate figures: the holdout half of the blind rows, 2,259 rows that took no part in choosing the winner
Balanced-accuracy lift: 20.3% (90% range 16.0-25.1%)
Balanced accuracy: 60.2% vs 50.0% for the baseline
Baseline: always predict the most common training class ('no')

Compare figures: all 4,518 blind rows, including the half the winner was chosen on
Correct: 4,029 of 4,518
Confusion (rows = actual, columns = predicted): actual 0 (no): 0 (no) 3,910, 1 (yes) 82; actual 1 (yes): 0 (no) 407, 1 (yes) 119
The product marks these Compare figures as diagnostic: "These results do not establish predictive performance" (outcome timing was not declared for this run).

The predictions file
4,518 held-out rows, one per line: heldout_row, age, job, marital, education, default, balance, housing, loan, contact, day, month, campaign, pdays, previous, poutcome, actual_y, predicted_y, confidence.
heldout_row numbers the held-out rows 1 to N in the order the product tested them. It is not a row number from the original dataset: the split shuffled the rows.

Run: model 2d9da143-a5fe-4508-b466-d8e915f354ac, tournament bank_full_nodur_20261028-212728, completed 2026-10-28T21:31:02Z
Dataset file sha256: a78d2754fdc337f204cd36a15f46d611ec33211f4e8ef90ece3798221954daab