Checklist Moldova AI Olympiad 2026 Simulation #1 (selection round) · Task 2
The Golden Card of the Codru
Predict a real-valued loyalty score for payment cards from merged card and transaction features, after two counting subtasks.
The task
A village marketplace wants to know which payment cards will bring long-term value. The data describe cards and their past interactions with merchants; card information, historical transactions, new merchant transactions and merchant details have already been merged into card-level tables with a real-valued loyalty score (target).
Subtask 1 (10 pts) counts training cards in four target bands (< −10, −10 to < 0, 0 to < 5, ≥ 5). Subtask 2 (20 pts) counts training cards in four bands of hist_transactions_count (≤ 10, 11–50, 51–100, > 100). Subtask 3 (70 pts) predicts the loyalty score of every test card.
Abridged by SOTA from the official materials. The official statement has the exact rules, and it wins wherever this summary differs.
At a glance
- You get
train.csv(card-level features with target),test.csv(same features without target),sample_submission.csv.- You submit
- One CSV with subtaskID, datapointID, answer: group names and counts (subtasks 1 and 2); card identifier and predicted score (subtask 3).
- Scoring
- Subtasks 1 and 2: exact counts. Subtask 3: RMSE; 0 pts if RMSE > 4, linear between 10 and 69 pts for 3.665 < RMSE ≤ 4, 70 pts if RMSE ≤ 3.665.
- Rules
- Python, Google Colab, Kaggle Notebook and standard data-science/ML libraries are allowed.
- No collaboration and no help from other people or AI assistants; internet only for documentation.
- Zoom session with screen sharing, camera, microphone and phone camera, plus OBS screen recording, for the whole round.
- Submission limit 50 (platform settings).
- Format
- Simulation #1, a preparatory round of the National AI Olympiad, 16 May 2026, online on Nitro AI Judge (3 hours from 10:00 local time; platform window 07:00–10:25 UTC), proctored via Zoom and OBS; its results count 30% towards selecting the 12 finalist teams.