Here is the formula and algorithm for computing kappa statistics

This test is used for testing the agreement between two different process of judgements. The closer Kappa is to 1.0, the higher the accuracy of the data.

Classification on Ground

Data base A B C D E Total

A: Live Oak 2 0 2 0 0 4 D (Diagonal total)= 12

B: Grass 0 1 0 0 0 1 q = x.i * x.i / N

C: Red Oak 1 0 1 0 0 2 q = 4*3/16+1*1/16+

D: Willow 0 0 0 2 1 3 2*3/16+3*2/16+

E: Water 0 0 0 0 6 6 6*7/16 = 4.1875

Total 3 1 3 2 7 N=16 Kappa = (D-q) / (N-q)

Pasrte it to notepad for easier understanding

Thank you so much Dear and respected Murli Mohan khetan, can u share any reference with me in which these formulas are reported, once again thank you