What is MI (Mutual Information)?

The Mutual information (MI) of two random variables is a proportion of the common reliance between the two factors. It is a dimensionless amount with (generally) units of bits and can be considered as the decrease in vulnerability around one irregular variable given information on another. Mutual information is utilized in deciding the closeness of two distinct clusterings of a dataset. In that capacity, it gives a few benefits over the traditional Rand index. Mutual information is utilized in clinical imaging for picture enlistment.