Home / experiments / data_arrays
Name Modified Size InfoDownloads / Week
Parent folder
svg_X_full_bal_True.npy 2016-07-07 3.4 MB
0
svg_Y_bal_False.npy 2016-07-07 904 Bytes
0
svg_Y_bal_True.npy 2016-07-07 832 Bytes
0
svg_X_reduced_bal_True.npy 2016-07-07 687.4 kB
0
svg_X_reduced_bal_False.npy 2016-07-07 804.3 kB
0
svg_X_full_bal_False.npy 2016-07-07 3.7 MB
0
ibd_X_full_bal_False.npy 2016-07-07 5.0 MB
0
svg_reduced_indices_bal_True.npy 2016-07-07 3.7 kB
0
svg_reduced_indices_bal_False.npy 2016-07-07 4.0 kB
0
svg_corMat.npy 2016-07-07 6.7 MB
0
svg_cooccurences.npy 2016-07-07 5.0 MB
0
svg_bactNames_reduced_bal_True.npy 2016-07-07 17.4 kB
0
svg_bactNames_reduced_bal_False.npy 2016-07-07 18.6 kB
0
svg_AAnumbers_bal_False.npy 2016-07-07 801 Bytes
0
svg_AAnumbers_bal_True.npy 2016-07-07 738 Bytes
0
mva_svg_species.npy 2016-07-07 760 Bytes
0
mva_svg_peaks.npy 2016-07-07 952 Bytes
0
mva_ibd_species.npy 2016-07-07 640 Bytes
0
ibd_Y_bal_True.npy 2016-07-07 976 Bytes
0
mva_ibd_peaks.npy 2016-07-07 1.0 kB
0
ibd_Y_bal_False.npy 2016-07-07 1.2 kB
0
ibd_X_reduced_bal_True.npy 2016-07-07 1.1 MB
0
ibd_X_reduced_bal_False.npy 2016-07-07 1.3 MB
0
ibd_X_full_bal_True.npy 2016-07-07 4.0 MB
0
ibd_corMat.npy 2016-07-07 11.5 MB
0
ibd_reduced_indices_bal_False.npy 2016-07-07 4.9 kB
0
ibd_reduced_indices_bal_True.npy 2016-07-07 4.9 kB
0
ibd_cooccurences.npy 2016-07-07 8.6 MB
0
ibd_bactNames_reduced_bal_True.npy 2016-07-07 22.9 kB
0
ibd_AAnumbers_bal_True.npy 2016-07-07 864 Bytes
0
ibd_bactNames_reduced_bal_False.npy 2016-07-07 22.9 kB
0
ibd_AAnumbers_bal_False.npy 2016-07-07 1.1 kB
0
Totals: 32 Items   52.0 MB 0 
This repository holds the implementation for the paper "Explaining individual classifier decisions for microbiota diagnosis", by A. Eck, L.M. Zintgraf, E.F.J. de Groot, T.G.J. de Meij, T.S. Cohen, P.H.M. Savelkoul, M. Welling, and A.E. Budding.

The important class is in prediction_difference_analysis.py, which handles all the computations. For a given input vector to a probabilistic classifier, an according relevance vector can be calculated by calling univ_rel().

The script run_experiment.py can be executed to run a simple experiment, returning for each sample in the dataset a relevance vector. Settings can be adjusted in the script (like which classifier to use).
Source: README.txt, updated 2016-07-07