Gallery

In this page, we demonstrate the output of various oversampling and noise removal techniques, using default parameters.

For binary oversampling and nosie removal, an artificial database was used, available in the datasets module.

Oversampling sample results

In the captions of the images some abbreviations referring to the operating principles are placed. Namely:

NR: noise removal is involved

DR: dimension reduction is applied

Clas: some supervised classifier is used

SCmp: sampling is carried out componentwise (attributewise)

SCpy: sampling is carried out by copying instances

SO: ordinary sampling (just like in SMOTE)

M: memetic optimization is used

DE: density estimation is used

DB: density based - the sampling is based on a density of importance assigned to the instances

Ex: the sampling is extensive - samples are added successively, not optimizing the holistic distribution of a given number of samples

CM: changes majority - even majority samples can change

Clus: uses some clustering technique

BL: identifies and samples the neighborhoods of borderline samples

A: developed for a specific application

_images/NoSMOTE.png

_images/SMOTE.png

_images/SMOTE_TomekLinks.png

_images/SMOTE_ENN.png

_images/Borderline_SMOTE1.png

_images/Borderline_SMOTE2.png

_images/ADASYN.png

_images/AHC.png

_images/LLE_SMOTE.png

_images/distance_SMOTE.png

_images/SMMO.png

_images/polynom_fit_SMOTE_bus.png

_images/polynom_fit_SMOTE_mesh.png

_images/polynom_fit_SMOTE_poly.png

_images/polynom_fit_SMOTE_star.png

_images/Stefanowski.png

_images/Safe_Level_SMOTE.png

_images/MSMOTE.png

_images/DE_oversampling.png

_images/SMOBD.png

_images/SUNDO.png

_images/MSYN.png

_images/SVM_balance.png

_images/TRIM_SMOTE.png

_images/SMOTE_RSB.png

_images/ProWSyn.png

_images/SL_graph_SMOTE.png

_images/NRSBoundary_SMOTE.png

_images/LVQ_SMOTE.png

_images/SOI_CJ.png

_images/ROSE.png

_images/SMOTE_OUT.png

_images/SMOTE_Cosine.png

_images/Selected_SMOTE.png

_images/LN_SMOTE.png

_images/MWMOTE.png

_images/PDFOS.png

_images/IPADE_ID.png

_images/RWO_sampling.png

_images/NEATER.png

_images/DEAGO.png

_images/Gazzah.png

_images/MCT.png

_images/ADG.png

_images/SMOTE_IPF.png

_images/KernelADASYN.png

_images/MOT2LD.png

_images/V_SYNTH.png

_images/OUPS.png

_images/SMOTE_D.png

_images/SMOTE_PSO.png

_images/CURE_SMOTE.png

_images/SOMO.png

_images/CE_SMOTE.png

_images/ISOMAP_Hybrid.png

_images/Edge_Det_SMOTE.png

_images/CBSO.png

_images/DBSMOTE.png

_images/ASMOBD.png

_images/Assembled_SMOTE.png

_images/SDSMOTE.png

_images/DSMOTE.png

_images/G_SMOTE.png

_images/NT_SMOTE.png

_images/Lee.png

_images/SPY.png

_images/SMOTE_PSOBAT.png

_images/MDO.png

_images/Random_SMOTE.png

_images/ISMOTE.png

_images/VIS_RST.png

_images/GASMOTE.png

_images/A_SUWO.png

_images/SMOTE_FRST_2T.png

_images/AND_SMOTE.png

_images/NRAS.png

_images/AMSCO.png

_images/SSO.png

_images/DSRBF.png

_images/NDO_sampling.png

_images/Gaussian_SMOTE.png

_images/kmeans_SMOTE.png

_images/Supervised_SMOTE.png

_images/SN_SMOTE.png

_images/CCR.png

_images/ANS.png

_images/cluster_SMOTE.png

_images/E_SMOTE.png

_images/ADOMS.png

_images/SYMPROD.png

Noise removal sample results

_images/TomekLinkRemoval.png

_images/CondensedNearestNeighbors.png

_images/OneSidedSelection.png

_images/CNNTomekLinks.png

_images/NeighborhoodCleaningRule.png

_images/EditedNearestNeighbors.png

Multiclass sample results

_images/multiclass-SMOTE.png

_images/multiclass-Borderline_SMOTE1.png

_images/multiclass-Borderline_SMOTE2.png

_images/multiclass-ADASYN.png

_images/multiclass-distance_SMOTE.png

_images/multiclass-SMMO.png

_images/multiclass-polynom_fit_SMOTE_mesh.png

_images/multiclass-polynom_fit_SMOTE_bus.png

_images/multiclass-polynom_fit_SMOTE_star.png

_images/multiclass-polynom_fit_SMOTE_poly.png

_images/multiclass-Safe_Level_SMOTE.png

_images/multiclass-MSMOTE.png

_images/multiclass-SMOBD.png

_images/multiclass-MSYN.png

_images/multiclass-TRIM_SMOTE.png

_images/multiclass-SMOTE_RSB.png

_images/multiclass-ProWSyn.png

_images/multiclass-SL_graph_SMOTE.png

_images/multiclass-NRSBoundary_SMOTE.png

_images/multiclass-LVQ_SMOTE.png

_images/multiclass-SOI_CJ.png

_images/multiclass-ROSE.png

_images/multiclass-SMOTE_OUT.png

_images/multiclass-SMOTE_Cosine.png

_images/multiclass-Selected_SMOTE.png

_images/multiclass-LN_SMOTE.png

_images/multiclass-MWMOTE.png

_images/multiclass-PDFOS.png

_images/multiclass-RWO_sampling.png

_images/multiclass-DEAGO.png

_images/multiclass-MCT.png

_images/multiclass-ADG.png

_images/multiclass-KernelADASYN.png

_images/multiclass-MOT2LD.png

_images/multiclass-V_SYNTH.png

_images/multiclass-OUPS.png

_images/multiclass-SMOTE_D.png

_images/multiclass-CURE_SMOTE.png

_images/multiclass-SOMO.png

_images/multiclass-CE_SMOTE.png

_images/multiclass-Edge_Det_SMOTE.png

_images/multiclass-CBSO.png

_images/multiclass-DBSMOTE.png

_images/multiclass-ASMOBD.png

_images/multiclass-Assembled_SMOTE.png

_images/multiclass-SDSMOTE.png

_images/multiclass-G_SMOTE.png

_images/multiclass-NT_SMOTE.png

_images/multiclass-Lee.png

_images/multiclass-Random_SMOTE.png

_images/multiclass-A_SUWO.png

_images/multiclass-AND_SMOTE.png

_images/multiclass-NRAS.png

_images/multiclass-SSO.png

_images/multiclass-DSRBF.png

_images/multiclass-NDO_sampling.png

_images/multiclass-Gaussian_SMOTE.png

_images/multiclass-kmeans_SMOTE.png

_images/multiclass-Supervised_SMOTE.png

_images/multiclass-SN_SMOTE.png

_images/multiclass-CCR.png

_images/multiclass-ANS.png

_images/multiclass-cluster_SMOTE.png

_images/multiclass-ADOMS.png

_images/multiclass-SYMPROD.png

_images/multiclass-DE_oversampling.png