Jensen Shannon Divergence Calculator Online

Example Input X: 0.1, 0.2, 0.7 = These number sum must be upto 1 means 0.1 + 0.2 + 0.7 =1

Example Input Y: 0.2, 0.3, 0.5 = These number sum must be upto 1 means 0.2 + 0.3 + 0.5 = 1

Values must be in between 0 and 1

KL divergence does measure how similar two distributions are, but because it breaks both symmetry and the triangle inequality, it is not a metric. Since Jenson-Shannon divergence is a metric, it is sometimes referred to as Jenson-Shannon distance.

In high dimensional environments, such as generative adversarial networks, JS distance is utilised to address the problems of KL divergence. Empirically, the majority of high-dimensional real-world data resembles a low-dimensional manifold. Since they rarely coincide, it would be like trying to discover a needle in a haystack when maximising KL (Kullback leibler) divergence of model distribution between data distribution.

When distributions do not or barely cross, KL divergence becomes infinite and the gradient signal is zero. When there is little to no overlap, when p_θ(x)=0, JS divergence behaves better because it doesn't become infinite, but it still has the same issue of having zero gradient.

How to use Calculate Jensen Shannon divergence using Calculator?

• You need to enter the value of X - Probabilities in the first input field given in the form. Remember each probability is seperated with comma and each value must be in between 0 and 1.
• Then enter the value of Y - Probabilities in the second input field given in the form. Remember each probability is seperated with comma and each value must be in between 0 and 1.
• Click on the button submit, and the results will automatically appear below in the result section.

More Calculation Examples (Jensen Shannon Divergence Examples)

• Example 1
• Probabilities - X : 0.1, 0.3, 0.6
• Probabilities - Y : 0.1, 0.2, 0.7
• Result : 0.006958856339414


• Example 2
• Probabilities - X : 0.2, 0.3, 0.5
• Probabilities - Y : 0.6, 0.3, 0.1
• Result : 0.12510060588455


• Example 3
• Probabilities - X : 0.5, 0.2, 0.3
• Probabilities - Y : 0.2, 0.3, 0.5
• Result : 0.050874612539598