ITU-ML5G-PS-032-KDDI-UT-NakaoLab-AI

Check the Python version：.
Check the Python version： $ python3 -V
Switch to project directory: $ cd <Your path>/itu-ml-challenge
If you have not completed the extraction steps: Modify the path information in the configuration file. Make sure <TRAIN_PATH>, <TEST_PATH> (conf.yaml) these two folders exist. If not, create it with mkdir. Then run: (`Attention: This step will take about 10+ hours to extract all the JSON files into ./dataset/csv-for-learning and ./dataset/csv-for-evluation folders. We have already checked in the above two folders so that we can skip this step. $ python3 1_feature_extract.py
Check that CSV files have been generated under <TRAIN_PATH> and <TEST_PATH>. Then run: $ python3 2_feature_combine.py
Check whether dataset.csv and testset.csv have been generated under ./csv/ in the current directory. Then run: $ python3 3_feature_refine.py
Check whether diff_dataset.csv and diff_testset.csv have been generated under ./csv/ in the current directory. If these two files have been generated, congratulations on the completion of the extraction. Then run: $ python3 4_train.py
You will see the training and test results printed on the console. At the same time, you can also use jupyter notebook to analyze the data. jupyter notebook