Submissions are scored on the **root mean squared error**. RMSE is defined as:

$$ \text{RMSE} = \sqrt{\frac{1}{n} \sum\_{i=1}^{n} (y\_i - \hat{y}\_i)^2} $$

where \\( \hat{y} \\) is the predicted value, \\( y \\) is the ground truth value, and \\( n \\) is the number of rows in the test data.

## Submission File
For each row `id` in the test set, you must predict the value of the target `loss` as described on the data tab, each on a separate row in the submission file. The file should contain a header and have the following format:

    id,loss
    250000,0.0
    250001,10.3
    250002,42.42
    etc.