ML5G-PS-012-Xdding

Team Xdding's solution to the Home User Network Classification problem statement by ZTE.

Competition Data Zip Google Drive Link

https://drive.google.com/file/d/1LUT96tVeihO8YIZAPBPL3suLUFj4H2W3/view?usp=sharing

Repository Overview

models: Directory storing the best RNN checkpoint and TSFresh + PCA + XGBoost model.

notebooks: Directory storing notebooks for each approach mentioned in the report: ROCKET classifier, LSTM RNN, manual feature extraction + XGBoost model, TSFresh + PCA + XGBoost model, and ts_regularization.ipynb to generate preprocessed data for ROCKET/LSTM RNN notebooks from included zip file (please download from Google Drive link above and unzip first).

ML5G-PS-012-Xdding-Presentation.pptx: Project presentation.

ML5G-PS-012-Xdding-Report.pdf: Project write-up.

requirements.txt: Packages that need to be installed if you want to run the notebooks. GPU is highly recommended for LSTM RNN notebook.

README.md: You are here!