The evaluation metric for this competition is [Mean Absolute Error or MAE](https://en.wikipedia.org/wiki/Mean_absolute_error). The Mean Absolute Error is a measure of errors between paired observations expressing the same phenomenon.

<img 
    style="display: block; 
           margin-left: auto;
           margin-right: auto;
           width: 30%;"
    src="https://www.gstatic.com/education/formulas2/553212783/en/mean_absolute_error.svg" 
    alt="MAE">
</img>

where:
- `y_i` is the true value of the i-th observation.
- `x_i` is the predicted value of the i-th observation.
- `n` is the total number of observations in the dataset.

# Submission Format
**For every row (product-location day combination) in the dataset,** submission files should contain two columns: `Id` and `Predicted`.`Id` is the string concatenation (use '_') of the `P_ID`, `L_ID` and `DATE` entry of the respective row. `Predicted` will contain the predicted demand for that row based on your ML model. For example, the forecast of '8' for the product '118' in location '14' on the '2021-12-01' will be expressed in the following format:
```
Id,Predicted
118_14_2021-12-01,8
```
The products and locations in the evaluation period are divided into two subsets: public and private scores.
The public score will be visible immediately after the submission and you can get your beverage at the boost after your first submission.
The private score will be revealed on Tuesday to define the winners for additional prizes.

Also, the evaluation period has cleaned and improved sales data, while training period includes a number of data problems