Refer to this journal '[A Survey on Trust Modeling][1]'-ACM computing survey,equation no:17
The equation is related to calculate uncertainty from probability:
H(p) = −plog2(p) − (1 − p) log2(1 − p)
So according to this equation when probability, p=0.5, then uncertainty is maximum.
Now I want to calculate uncertainty from the correlation. I have two array which outputs the probability of a dress to be red. I calculate correlation(value range -1 to 1) between these two array. So, when the correlation is greater than zero, so both array agree about the red color, if zero-it follows no relationship and for less than zero-one array agrees and other disagree about the red color. So when the value is zero-maximum uncertainty.So my problem is to calculate the uncertainty from these correlation value. I want to find an equation similar to the above equation that supports uncertainty behavior from correlation. The frame is=(red,not red,red or not red, none). 'red or not red' is actually uncertainty part.
[1]: Jin-Hee Cho, Kevin Chan, and Sibel Adali. 2015. A Survey on Trust Modeling. ACM Comput. Surv. 48, 2, Article 28 (November 2015), 40 pages. https://doi.org/10.1145/2815595
