This is the code which has been used to run the experiments of the project Machine Learning in Heart Disease Prediction - Analyzing the Influence of Small Sample Sizes which was part of the small sample sie project seminar at HPI.
The following files are scripts belonging to the experiments
- ml_performance.py is the script used to execute the first experiment of section \ref{sec:mlp_lows}.
- overfitting_test.py is the script used to execute the second experiment of section \ref{sec:mlp_lows} which dealt with the overfitting.
- subgroups_ml_performance.py is the script used to execute the first experiment of section \ref{sec:mlp_lows_subs}.
- subgroups_overfitting_test.py is the script used to execute the second experiment of section \ref{sec:mlp_lows_sub} which dealt with the overfitting.
- feature_selection.py is used to run the experiment from section \ref{sec:feature_selection}.
Please be aware, that the sampling of the data is a random process. Thus there is no guarantee that executing the scripts will yield identical results.
The repository also includes the initial data analysis and the analysis of the results.