Skip to content

Experiments conducted on the TPEHGDB dataset to reproduce the reported results from "A critical look at studies applying over-sampling on the TPEHGDB dataset"

License

Notifications You must be signed in to change notification settings

fongenae/TPEHGDB-Experiments

 
 

Repository files navigation

Experiments on the TPEHGDB dataset

The impact of features extracted from raw EHG signals

To compare the predictive performances of the model constructed using solely clinical variables and the model where the data is augmented with four features, extracted from the EHG signal, run python baseline_clinical.py and python baseline_ehg.py:

The impact of features extracted from EHG signals

Oversampling on an artificial dataset

To generate the different scatter plots, used for Figure 1 (see below), run python oversampling.py

Oversamping correctly vs incorrectly (artificial dataset)

Oversampling the TPEHGDB dataset incorrectly vs correctly

To compare oversampling in a correct fashion vs oversampling in an incorrect fashion, run python baseline_smote_correct.py and python baseline_smote_incorrect.py:

Oversamping correctly vs incorrectly (TPEHGDB dataset)

Questions, citing and contact

A citation will be posted here as soon as this is available. In case there are any problems with running the code, or any questions, please create an issue or contact me at gilles(dot)vandewiele(at)ugent(dot)be

About

Experiments conducted on the TPEHGDB dataset to reproduce the reported results from "A critical look at studies applying over-sampling on the TPEHGDB dataset"

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%