Jensen-Shannon Divergence (JS) - Amazon SageMaker
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

Jensen-Shannon Divergence (JS)

The Jensen-Shannon divergence (JS) measures how much the label distributions of different facets diverge from each other entropically. It is based on the Kullback-Leibler divergence, but it is symmetric.

The formula for the Jensen-Shannon divergence is as follows:

        JS = ½*[KL(Pa || P) + KL(Pd || P)]

Where P = ½( Pa + Pd ), the average label distribution across facets a and d.

The range of JS values for binary, multicategory, continuous outcomes is [0, ln(2)).

  • Values near zero mean the labels are similarly distributed.

  • Positive values mean the label distributions diverge, the more positive the larger the divergence.

This metric indicates whether there is a big divergence in one of the labels across facets.