## 4. Submission

Submit a CSV with columns `Id` and `result`:

```python
import pandas as pd

submission = pd.DataFrame({
    'Id': test_df['Id'],
    'result': predictions
})
submission.to_csv('submission.csv', index=False)
```

## 5. Score

Accuracy: fraction of correct assignments.

$$ \text{Accuracy} = \frac{\text{Number of Correct Predictions}}{70} $$

- **Baseline Score:** ~0.08
- **Reference Solution Score:** ~0.41

The baseline can be found on the **[Code](https://www.kaggle.com/code/antoningorokva/baseline-alchemy-aicc-round-4)** section of this competition or in Google Colab [Here](https://colab.research.google.com/#fileId=https%3A//storage.googleapis.com/kaggle-colab-exported-notebooks/antoningorokva/baseline-alchemy-aicc-round-4.fedbf9a4-dcae-41e7-8d23-4fd53d64eb73.ipynb%3FX-Goog-Algorithm%3DGOOG4-RSA-SHA256%26X-Goog-Credential%3Dgcp-kaggle-com%2540kaggle-161607.iam.gserviceaccount.com/20260219/auto/storage/goog4_request%26X-Goog-Date%3D20260219T205732Z%26X-Goog-Expires%3D259200%26X-Goog-SignedHeaders%3Dhost%26X-Goog-Signature%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)