Table 55 Features by supervised ranking feature importance
From: IoT information theft prediction using ensemble feature selection
| XGBoost | CatBoost | Light GBM | RF |
|---|---|---|---|
| dport | dport | dport | state_number |
| state_number | state_number | dur | dport |
| stddev | bytes | flgs_number | bytes |
| pkts | dur | proto_number | dur |
| proto_number | max | drate | flgs_number |
| sum | proto_number | state_number | rate |
| dur | dbytes | sport | sbytes |
| dbytes | mean | rate | dbytes |
| flgs_number | sport | srate | dpkts |
| mean | pkts | mean | sum |
| rate | sbytes | drate | |
| dpkts | min | min | |
| sport | sum | proto_number | |
| max | dbytes | sport | |
| spkts | pkts | srate | |
| drate | stddev | mean | |
| srate | bytes | spkts | |
| min | spkts | max | |
| dpkts | |||
| max |