Submissions are evaluated using the **Symmetric Mean Absolute Percentage Error (SMAPE)** between the predicted and actual demand values.

The SMAPE metric measures the accuracy of your forecasts as a percentage and is defined as:

$$
SMAPE = \frac{100%}{N} \sum_{i=1}^{N} \frac{|y_{pred,i} - y_{true,i}|}{(|y_{true,i}| + |y_{pred,i}|)/2}
$$

A **lower SMAPE** indicates a more accurate forecast.

---

## Submission File

For each date in the test set, you must predict the expected **bus demand (`target`)**.
The file must include a header and follow this format:

```
date,target
2021-03-25,0.0
2021-03-26,23.0
2021-03-27,567.0
2021-03-28,644.0
2021-03-29,1182.0
2021-03-30,1345.0
2021-03-31,2547.0
2021-04-01,2465.0
2021-04-02,2175.0
2021-04-03,2258.0
```

Each row should contain the **forecasted demand** for the corresponding **date** in the test dataset.