README.md

metadata

license: apache-2.0
language:
  - 'no'
  - nb
  - nn
pipeline_tag: text-classification
library_name: fasttext

Test results

Binary classification accuracy (threshold at label 3) ≈ 74.73%

Labels: [0, 1, 2, 3, 4, 5] Confusion Matrix:

 [[62 30  2  6  0  0]
 [24 43 14 15  4  0]
 [18 37 22 23  0  0]
 [ 7 25 17 43  8  0]
 [ 5 15 13 47 19  1]
 [ 4  3  2 26 14  1]]

Classification Report:

              precision    recall  f1-score   support

           0       0.52      0.62      0.56       100
           1       0.28      0.43      0.34       100
           2       0.31      0.22      0.26       100
           3       0.27      0.43      0.33       100
           4       0.42      0.19      0.26       100
           5       0.50      0.02      0.04        50

    accuracy                           0.35       550
   macro avg       0.38      0.32      0.30       550
weighted avg       0.37      0.35      0.32       550