The evaluation metric for this competition is classification accuracy, which is computed as follows:
\[ Acc = 1-Err_{test} = 1-\frac{1}{n_{test}}\sum_{i=1}^{n_{test}} |f_{\theta}(\textbf{x}_i)-y_i| \]
where: \\((\textbf{x}_i,y_i)\\) is a test example, \\(n_{test}\\) is the total number of test examples, and \\(f_{\theta}\\) is the prediction function you have learned.
Submission Format
For every team in the competition
, submission files should contain exactly two columns: Id and Category. Id is a unique ID per datapoint, while Category indicates whether the example is an instance of cognates (value: 1) or not (value: 0). For example:
Id,Category
540102,0