Discord

Checklist OPEIA 2025 Competition · Task 1

The Mistic Alpaca

Engineer at most 100 features from 32×32 textile fragments so that a plain linear regression predicts each textile's value.

  • Tabular
  • Feature engineering for regression

The task

In an Andean legend, the Mistic Alpaca Qoriñawi weaves textiles whose value must be predicted; only a 32×32 corner fragment of each textile is available. Contestants may use only scikit-learn LinearRegression with default hyperparameters and at most 100 engineered features; feature generation, training and inference must take no more than 60 seconds on a standard non-GPU Colab instance.

Supervised models may not be used as feature extractors (simple models may be used for feature selection); unsupervised learning is allowed; pretrained models and AutoML are not; validation data may not be used for training.

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
A pickle file downloaded in the notebook (1_ml_data_start.pickle) with training and validation data.
You submit
final_predictions.csv and the completed notebook.
Scoring
Root Mean Square Error.
Rules
  • LinearRegression only; ≤ 100 features; ≤ 60 s per model on CPU.
Format
OPEIA 2025, a 14-day team competition (teams of four listed on the results page); dates not stated on the pages opened.

Details

Year
2025, Online
Round
Competition · Task 1
Language
English
License
Not stated by the source