Tversky index

Measuring the similarity between sets is useful in many application. The Jaccard index is probably one of the most well known set similarity indexes. The Tversky index is a more general index, which includes the Jaccard or Dice indexes as a special case when \(\gamma\) and \(\eta\) are both equal to 1, or both equal to 0.5, respectively.

\[sim(X,Y) = \frac{|X \cap Y|}{|X \cap Y| + \gamma|X - Y| + \eta |Y - X|}\]

(see http://en.wikipedia.org/wiki/Tversky_index)