- Here constains：Training model, Pre-processed code, Training and Testing code
- NAD2021 challenge official website : https://nad2021.nctu.edu.tw/
- Syslog traffics dataset from ZYELL
@2021 IEEE International Conference on Acoustics, Speech and Signal Processing 6-11 June 2021 • Toronto, Ontario, Canada
- Arthor ： Shao-Wen, Cheng
| File NAME | decription |
|---|---|
| Preprocessed code | contain three python files, for some featrues processing and data processing. |
| Testing code | Contain eight folders, version 2 obtain better scrore. Each folders contains two python files for loading testing datasets and fitting to get predictions.Models used here are contain in Training model.rar, see detailed info below |
| Training code | Contain six python files. s1_training_model_flow.py is for training model flow (model algorithm can choose: Random Forest, XGBoost) score.py is score calculating function based on the calculate method provided by official to calculate the score of this model. |
| Training model | models use update in "Realease" |
Due to the license agreement signed with official, so i won't update either training datasets or testing datasets.
If there's ant problem, please contact me:
e-mail：shaowen.eic09g@nctu.nctu.edu