Table 8.
Single-label classification: performance metrics of RF classifier
single-label classification: | ||||
---|---|---|---|---|
best performing GB classifier | ||||
labels | precision | recall | F1-score | support |
helper_functions | 0.82 | 0.79 | 0.81 | 154 |
load_data | 0.76 | 0.70 | 0.73 | 158 |
data_preprocessing | 0.74 | 0.77 | 0.75 | 591 |
data_exploration | 0.63 | 0.81 | 0.71 | 432 |
modelling | 0.69 | 0.65 | 0.67 | 316 |
evaluation | 0.49 | 0.28 | 0.36 | 95 |
prediction | 0.55 | 0.20 | 0.29 | 60 |
result_visualization | 0.88 | 0.50 | 0.64 | 30 |
save_results | 0.82 | 0.51 | 0.63 | 45 |
comment_only | 1.00 | 0.86 | 0.93 | 37 |
accuracy | 0.71 | 1918 | ||
macro avg | 0.74 | 0.61 | 0.65 | 1918 |
weighted avg | 0.71 | 0.71 | 0.70 | 1918 |