Gallery Of Evaluation Metric

Selection of suitable evaluation metric plays a significant role in learning best mathmetic model, as well as offering a uniformed approach to measure the performance of the data mining models from different perspectives. This post aims to set up a collection for existing evaluation metrics, which will be organized by the data mining tasks, such as top-k ranking, rating prediction, stability, diversity etc.

Ranking

“Precision#K”

“Recall#K”

“nDCG”

Classification

Receiver operating characteristic (ROC) is a graphical plot illustrating the performance of a binary classifier system by plotting the fraction of true positives out of total actual positives (TPR = true positive rate) vs. the fraction of false positives out of total actual negatives (FPR = false positive rate).

Ref. Wikipedia