Table 10.
Multi-label classification: performance metrics of RF classifier
multi-label classification: | ||||
---|---|---|---|---|
best performing RF classifier | ||||
labels | precision | recall | F1-score | support |
helper_functions | 0.91 | 0.82 | 0.86 | 286 |
load_data | 0.88 | 0.66 | 0.75 | 191 |
data_preprocessing | 0.69 | 0.75 | 0.72 | 437 |
data_exploration | 0.80 | 0.79 | 0.79 | 801 |
modelling | 0.78 | 0.58 | 0.67 | 361 |
evaluation | 0.74 | 0.30 | 0.43 | 171 |
prediction | 0.77 | 0.33 | 0.47 | 90 |
result_visualization | 0.43 | 0.16 | 0.24 | 97 |
save_results | 0.84 | 0.53 | 0.65 | 30 |
comment_only | 1.00 | 0.86 | 0.93 | 37 |
micro avg | 0.78 | 0.67 | 0.72 | 2501 |
macro avg | 0.78 | 0.58 | 0.65 | 2501 |
weighted avg | 0.78 | 0.67 | 0.71 | 2501 |
samples avg | 0.74 | 0.71 | 0.71 | 2501 |