**Evaluation method**
To evaluate the results, you will send only one .csv file each time, in the format below. Additionally, during the competition, you will be able to send a maximum of 2 solutions per day.

In the leaderboard, you will see the ranking based on 40% of the final evaluation dataset. The score for the other 60% of the data set will be hidden and shown after the competition ends.

In the end, you will be able to choose two solutions to be evaluated on the final data set.


**Submission Format**
For each movie in the dataset, the submission file must contain two columns: movie_id and popularity.

The uploaded file must be saved with the .csv extension in the following format:

<pre>title_index, popularity
0, popular
1, unpopular
2, popular
</pre>

**Evaluation metrics**

For evaluation will be used the ["Classification accuracy"](https://en.wikipedia.org/wiki/Accuracy_and_precision) metric.

The classification accuracy defines the number of correctly classified examples divided by the total number of examples in the dataset.