For evaluating the models, *accuracy* of predictions and *innovation* of models play roles. Provided that you want your model to be evaluated at FOMLIG2025, upload *at least* a two-page summary of your developed model, to present the innovation also. Though, single-page graphical abstracts are more than welcome to be presented at FOMLIG workshop in Florence, Italy, 15-17 October. If not, you can evaluate your model here on Kaggle.

Submissions are evaluated on [mean squared error (MSE)](https://en.wikipedia.org/wiki/Mean_squared_error) between the predicted values and the observed V<sub>s</sub>.

The *winners*:
- One winner from Kaggle (based on the Kaggle Leaderboard ranking),
- Two winners from FOMLIG participants (based on the accuracy and innovation criteria evaluated at FOMLIG),
- One winner from open participants who may not have the opportunity to join us in Italy (based on the accuracy and innovation criteria evaluated at FOMLIG),

will be awarded Certificates.

## Submission File
For each ID in the test set, you must predict V<sub>s</sub> variable. The file should contain a header and have the following format, as shown in 'Sample submission.csv':

    Project,Vs
    Test_1,0
    Test_2,0
    Test_3,0
    etc.