Submissions will be evaluated using the root mean square ([RMSE](https://en.wikipedia.org/wiki/Root_mean_square)) metric, which measures how closely the predicted preservation_score values align with the actual scores. A lower RMSE indicates more accurate predictions. While RMSE is the official evaluation metric, participants may also consider tracking the R² score during development to assess how much variance in preservation can be explained by their models. To improve performance, participants are encouraged to experiment with feature engineering, multimodal fusion (text and image), and a variety of machine learning techniques.

## Submission File
For each id in the test set, you must predict a probability for the preservation_score variable. The file should contain a header and have the following format:

```
id,preservation_score
8001,74.59488588167487
8002,74.59488588167487
8003,74.59488588167487
8004,74.59488588167487
8005,74.59488588167487
8006,74.59488588167487
8007,74.59488588167487
8008,74.59488588167487
8009,74.59488588167487
8010,74.59488588167487
```