Learning Data Mining with Python - Second Edition

Other Evaluation Metrics

There is a wide range of evaluation metrics for other takes. Some notable ones to investigate are:

• The Lift Metric: https://en.wikipedia.org/wiki/Lift_(data_mining)
• Segment evaluation metrics: http://segeval.readthedocs.io/en/latest/
• Pearson's Correlation Coefficient: https://en.wikipedia.org/wiki/Pearson_correlation_coefficient
• Area under the ROC Curve: http://gim.unmc.edu/dxtests/roc3.htm
• Normalized Mutual Information: http://scikit-learn.org/stable/modules/clustering.html#mutual-info-score

Each of these metrics was developed with a particular application in mind. For example, the segment evaluation metrics evaluate how accurate breaking a document of text into chunks is, allowing for some variation between chunk ...