For processing the ECG data, some libraries were used, namely: PyWT, Scipy and Biosppy. The processing code is available in the process_dataset.ipynb file.
The model used for data classification was a GBT (Gradient Boosted Trees) model called LightGBM. For its training, larger weights were used for the minority classes and the cross validation technique called Stratified KFold was also used.
The code for the model is in the model_f.ipynb file.
The final predictions selected on the platform are present in the output folder in the lgb_v2000_final.csv file.