## Evaluation

The evaluation of submissions in this competition will be based on how well the participant's predictions match the true values in the provided test dataset. Below are the specific evaluation criteria and how submissions will be scored.

### 1. Evaluation Metric

The primary evaluation metric for this competition will be the **F1 Score**. Accuracy is the fraction of correct predictions made by the model over the total number of predictions. The formula is:

Where:
- Correct predictions are the number of times the predicted class matches the true class in the test set.
- Total number of predictions is the total count of predictions made for the test data.

The model with the highest accuracy will be considered the winner. During the competition, accuracy will be calculated on the hidden test set and will be used to rank the submissions.

### 2. Submission Format

Participants must submit their predictions in a **CSV** file with the following structure:

- **Format**: The file must contain two columns: the first column should be the unique identifier for each sample in the test set, and the second column should contain the predicted class for each sample.
- **Structure**: The file should follow this structure, It must comply with the condition of using a comma (,) as the delimiter.

| ID | leaving |
|-----|------------|
| 1   | 0    |
| 2   | 1    |
| 3   | 1    |
| ... | ...        |

- **File Name**: The file should be named `predictions.csv`.

### 3. Restrictions

- **Submission Limit**: Participants can make a limited number of submissions throughout the competition. Please ensure that you test your model thoroughly before making your final submission.
- **Dataset Usage**: Predictions should be made using only the training and test data provided by the competition. The use of external datasets is not allowed.
- **Cross-validation**: We recommend using cross-validation techniques to assess the performance of your model during development, though final submissions will only be evaluated on the hidden test set.

### 4. Submission Evaluation

Each submission will be evaluated and ranked based on the **Accuracy Score** calculated on the test set. Evaluation results will be shown on the leaderboard, and the rankings will be updated in real-time as new submissions are received.

- **File Submission**: When a file is submitted, an automatic score will be generated based on the accuracy of the predictions. Participants can check this score and view their position on the leaderboard at any time during the competition.
- **Leaderboard**: The leaderboard will rank participants based on the highest accuracy achieved on the test set. Rankings will be updated each time a new submission is made.

### 5. Additional Considerations

- **Manual Review**: If suspicious behavior such as the use of unauthorized data or fraudulent submissions is detected, the organizers reserve the right to manually review the predictions and disqualify participants if necessary.
- **Final Results**: The winner of the competition will be the participant with the highest accuracy at the end of the competition, after the evaluation of predictions on the final test set.



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

    ID; Leaving
    2;0
    5;0
    6;0
    etc.