ABSTRACT
This chapter presents two widely accepted, basic analysis techniques Kappa and Margfit that can be performed once an error matrix has been properly generated. The Kappa analysis is a discrete multivariate technique used in accuracy assessment for statistically determining whether one error matrix is significantly different from another. The Kappa analysis could then be used to statistically determine whether the new algorithm is indeed the better approach. Consequently, the Kappa analysis has become a standard component of almost every accuracy assessment and is a standard component of most image analysis software packages that include accuracy assessment procedures. In addition to the Kappa analysis, a second technique called Margfit can be applied to “normalize” or standardize the error matrices for comparison purposes. Margfit uses an iterative proportional fitting procedure, which forces each row and column in the matrix to sum to a predetermined value; hence the name Margfit—marginal fitting.
