Learning multidimensional projections with neural networks
Mateus Espadoto · 2021
nonlinear distances global 0 $ (# ) no no yes Vispipeline GDA nonlinear distances global 1 $ (= 3 ) no no yes DR Toolbox GPLVM nonlinear distances global 1 $ (= 3 ) no no no DR Toolbox F-ICA linear samples global 2 $ (= 3 ) yes yes yes scikit-learn IDMAP nonlinear samples local 3 $ (# 2 ) no no yes Vispipeline ISO nonlinear samples local 1 $ (# 3 ) yes no yes scikit-learn L-ISO nonlinear samples local 1 $ (# 3 ) no no no Vispipeline LAMP nonlinear samples local 3 $ (# =) yes yes no Vispipeline LE nonlinear distances local 0 $ (# 3 ) no no no scikit-learn LLC nonlinear samples local 3 $ (8= 3 ) no no yes DR Toolbox LLE nonlinear samples local 3 $ (# 3 ) yes no no scikit-learn H-LLE nonlinear samples local 3 $ (# 3 ) yes no no scikit-learn M-LLE nonlinear samples local 3 $ (# 3 ) yes no no scikit-learn LMNN linear samples local 3 $ (= 2 ) no no yes DR Toolbox LPP linear samples global 1 $ (# 3 ) yes no yes Tapkee LSP nonlinear samples local 4 $ (# 3 ) no no yes Vispipeline LTSA nonlinear samples local 3