round2

ECG Multi-Label Classfication - Round 2

1. Structure

• code: scripts for preprocessing, training and testing;
• data: Round 2 raw data;
• user_data: data generated by user, including models;
• prediction_result: prediction output.

2. Run

Only test scripts in Ubuntu 18.04 and CentOS 7.6.

cd code
./main.sh

Three commands in main.sh:

# Preprocessing
python ./prep/ecg_prep.py -i ../data -o ../user_data

# Train model
python ./train/ecg_train.py \
    -l2 ../user_data/round2_train_weighted.csv \
    -l1 ../user_data/round1_merge_noDup_weighted.csv \
    -m ../user_data/models2_small \
    -b small -nt 6 -c 5 -g 0

# Test model
# Ensemble 5 predictions
python ./test/ecg_test.py \
    -s /tcdata/hf_round2_testB \
    -m ../user_data/models2 \
    -t /tcdata/hf_round2_subB.txt \
    -a ../data/hf_round2_arrythmia.txt \
    -o ../prediction_result \
    -b small -g 0

You can also run these three commands one by one.