Skip to main content

Table 4 Classification metrics for different models tested on Vojvodina dataset

From: Transfer learning approach based on satellite image time series for the crop classification problem

Model

Overall accuracy

Class-mean recall

Macro f-score

Weighted f-score

Kappa

BMP

0.924

0.767

0.764

0.923

0.889

FE

0.936

0.896

0.896

0.936

0.906

PFT

0.940

0.907

0.906

0.939

0.911

FFT

0.939

0.896

0.905

0.939

0.911

  1. Highest performing values are highlighted in bold