CNI - Challenge



Challenge Evaluation


Evaluation metrics



We utilize accuracy, error rate, sensitivity, specificity, precision, recall, F-Measure, Geometric-mean, AUC, and optimized precision.
(cf. Hossin and Sulaiman. A review on evaluation metrics for data classification evaluations. International Journal of Data Mining & Knowledge Management Process, 5(2):1,2015)

In order to do so, please ensure that your program creates two output files: one containing a hard classification (0 and 1; classification.txt), and a separate file with the prediction probability/score (scores.txt).

Justification of metrics



With the lack of consensus on which metric is most suitable to determine the most appropriate classifier, we will use an inclusive approach. This includes multiple measures which are commonly used in classification tasks, such as accuracy and AUC, allowing for a more intuitive interpretation of the results, in addition to measures such as Geometric-mean and optimized precision.
(cf. M Hossin and MN Sulaiman. A review on evaluation metrics for data classification evaluations. International Journal of Data Mining & Knowledge Management Process, 5(2):1,2015)