README.md

In bobdouma/kappa_multinomial: Evaluate the performance of probabilistic multinomial models

This is a brief tutorial for using kappa_multinomial, a new way to assess the performance of multinomial predictions for landcover

Increasingly, land cover models use multinomial models to predict occurence of discrete classes. At present, no suitable mesure exist to evaluate the agreement of such models with observations.

This package provides a method $k_multinomial$ to assess the agreement between multinomial models (i.e. those that predict an occurence probability for classes of outcomes) and observations which have either discrete outcomes or outcomes represented by frequencies.

We refer to Run_Kappa for examples and application of the above mentioned.

bobdouma/kappa_multinomial documentation built on May 12, 2019, 11:28 p.m.