• Dimension Estimation
  • Intrinsic dimension estimation algorithms try to estimate the rank/dimension of low-dimensional structure that is embedded in high-dimensional space.

    est.Ustat()

    ID Estimation with Convergence Rate of U-statistic on Manifold

    est.boxcount()

    Box-counting Dimension

    est.clustering()

    Intrinsic Dimension Estimation via Clustering

    est.correlation()

    Correlation Dimension

    est.danco()

    Intrinsic Dimensionality Estimation with DANCo

    est.gdistnn()

    Intrinsic Dimension Estimation based on Manifold Assumption and Graph Distance

    est.incisingball()

    Intrinsic Dimension Estimation with Incising Ball

    est.made()

    Manifold-Adaptive Dimension Estimation

    est.mindkl()

    MiNDkl

    est.mindml()

    MINDml

    est.mle1()

    Maximum Likelihood Esimation with Poisson Process

    est.mle2()

    Maximum Likelihood Esimation with Poisson Process and Bias Correction

    est.nearneighbor1()

    Intrinsic Dimension Estimation with Near-Neighbor Information

    est.nearneighbor2()

    Near-Neighbor Information with Bias Correction

    est.packing()

    Intrinsic Dimension Estimation using Packing Numbers

    est.pcathr()

    PCA Thresholding with Accumulated Variance

    est.twonn()

    Intrinsic Dimension Estimation by a Minimal Neighborhood Information

  • Linear Embedding
  • Although all linear methods are designed to find explicit projection matrix for embedding, we divide this part into two categories; feature selection to select a subset of variables to extract information at their own best measurements, and dimension reduction type to denote the rest.

    (2-1) Feature Selection

    do.cscore()

    Constraint Score

    do.cscoreg()

    Constraint Score using Spectral Graph

    do.disr()

    Diversity-Induced Self-Representation

    do.enet()

    Elastic Net Regularization

    do.fosmod()

    Forward Orthogonal Search by Maximizing the Overall Dependency

    do.fscore()

    Fisher Score

    do.lasso()

    Least Absolute Shrinkage and Selection Operator

    do.lscore()

    Laplacian Score

    do.lsdf()

    Locality Sensitive Discriminant Feature

    do.lsls()

    Locality Sensitive Laplacian Score

    do.lspe()

    Locality and Similarity Preserving Embedding

    do.mcfs()

    Multi-Cluster Feature Selection

    do.mifs()

    Mutual Information for Selecting Features

    do.nrsr()

    Non-convex Regularized Self-Representation

    do.pfa()

    Principal Feature Analysis

    do.procrustes()

    Feature Selection using PCA and Procrustes Analysis

    do.rsr()

    Regularized Self-Representation

    do.specs()

    Supervised Spectral Feature Selection

    do.specu()

    Unsupervised Spectral Feature Selection

    do.spufs()

    Structure Preserving Unsupervised Feature Selection

    do.udfs()

    Unsupervised Discriminative Features Selection

    do.ugfs()

    Unsupervised Graph-based Feature Selection

    do.uwdfs()

    Uncorrelated Worst-Case Discriminative Feature Selection

    do.wdfs()

    Worst-Case Discriminative Feature Selection

    (2-2) Linear Projection

    do.adr()

    Adaptive Dimension Reduction

    do.ammc()

    Adaptive Maximum Margin Criterion

    do.anmm()

    Average Neighborhood Margin Maximization

    do.asi()

    Adaptive Subspace Iteration

    do.bpca()

    Bayesian Principal Component Analysis

    do.cca()

    Canonical Correlation Analysis

    do.cnpe()

    Complete Neighborhood Preserving Embedding

    do.crp()

    Collaborative Representation-based Projection

    do.dagdne()

    Double-Adjacency Graphs-based Discriminant Neighborhood Embedding

    do.dne()

    Discriminant Neighborhood Embedding

    do.dspp()

    Discriminative Sparsity Preserving Projection

    do.elde()

    Exponential Local Discriminant Embedding

    do.elpp2()

    Enhanced Locality Preserving Projection (2013)

    do.eslpp()

    Extended Supervised Locality Preserving Projection

    do.extlpp()

    Extended Locality Preserving Projection

    do.fa()

    Exploratory Factor Analysis

    do.fssem()

    Feature Subset Selection using Expectation-Maximization

    do.ica()

    Independent Component Analysis

    do.isoproj()

    Isometric Projection

    do.kmvp()

    Kernel-Weighted Maximum Variance Projection

    do.kudp()

    Kernel-Weighted Unsupervised Discriminant Projection

    do.lda()

    Linear Discriminant Analysis

    do.ldakm()

    Combination of LDA and K-means

    do.lde()

    Local Discriminant Embedding

    do.ldp()

    Locally Discriminating Projection

    do.lea()

    Locally Linear Embedded Eigenspace Analysis

    do.lfda()

    Local Fisher Discriminant Analysis

    do.llp()

    Local Learning Projections

    do.lltsa()

    Linear Local Tangent Space Alignment

    do.lmds()

    Landmark Multidimensional Scaling

    do.lpca2006()

    Locally Principal Component Analysis by Yang et al. (2006)

    do.lpe()

    Locality Pursuit Embedding

    do.lpfda()

    Locality Preserving Fisher Discriminant Analysis

    do.lpmip()

    Locality-Preserved Maximum Information Projection

    do.lpp()

    Locality Preserving Projection

    do.lqmi()

    Linear Quadratic Mutual Information

    do.lsda()

    Locality Sensitive Discriminant Analysis

    do.lsir()

    Localized Sliced Inverse Regression

    do.lspp()

    Local Similarity Preserving Projection

    do.mds()

    (Classical) Multidimensional Scaling

    do.mfa()

    Marginal Fisher Analysis

    do.mlie()

    Maximal Local Interclass Embedding

    do.mmc()

    Maximum Margin Criterion

    do.mmp()

    Maximum Margin Projection

    do.mmsd()

    Multiple Maximum Scatter Difference

    do.modp()

    Modified Orthogonal Discriminant Projection

    do.msd()

    Maximum Scatter Difference

    do.mvp()

    Maximum Variance Projection

    do.nolpp()

    Nonnegative Orthogonal Locality Preserving Projection

    do.nonpp()

    Nonnegative Orthogonal Neighborhood Preserving Projections

    do.npca()

    Nonnegative Principal Component Analysis

    do.npe()

    Neighborhood Preserving Embedding

    do.odp()

    Orthogonal Discriminant Projection

    do.olda()

    Orthogonal Linear Discriminant Analysis

    do.olpp()

    Orthogonal Locality Preserving Projection

    do.onpp()

    Orthogonal Neighborhood Preserving Projections

    do.opls()

    Orthogonal Partial Least Squares

    do.pca()

    Principal Component Analysis

    do.pflpp()

    Parameter-Free Locality Preserving Projection

    do.pls()

    Partial Least Squares

    do.ppca()

    Probabilistic Principal Component Analysis

    do.rlda()

    Regularized Linear Discriminant Analysis

    do.rndproj()

    Random Projection

    do.rpcag()

    Robust Principal Component Analysis via Geometric Median

    do.rsir()

    Regularized Sliced Inverse Regression

    do.sammc()

    Semi-Supervised Adaptive Maximum Margin Criterion

    do.save()

    Sliced Average Variance Estimation

    do.sda()

    Semi-Supervised Discriminant Analysis

    do.sdlpp()

    Sample-Dependent Locality Preserving Projection

    do.sir()

    Sliced Inverse Regression

    do.slpe()

    Supervised Locality Pursuit Embedding

    do.slpp()

    Supervised Locality Preserving Projection

    do.spc()

    Supervised Principal Component Analysis

    do.spca()

    Sparse Principal Component Analysis

    do.spp()

    Sparsity Preserving Projection

    do.ssldp()

    Semi-Supervised Locally Discriminant Projection

    do.udp()

    Unsupervised Discriminant Projection

    do.ulda()

    Uncorrelated Linear Discriminant Analysis

  • Nonlinear Embedding
  • do.bmds()

    Bayesian Multidimensional Scaling

    do.cge()

    Constrained Graph Embedding

    do.cisomap()

    Conformal Isometric Feature Mapping

    do.crca()

    Curvilinear Component Analysis

    do.crda()

    Curvilinear Distance Analysis

    do.dm()

    Diffusion Maps

    do.dppca()

    Dual Probabilistic Principal Component Analysis

    do.dve()

    Distinguishing Variance Embedding

    do.fastmap()

    FastMap

    do.hydra()

    Hyperbolic Distance Recovery and Approximation

    do.idmap()

    Interactive Document Map

    do.iltsa()

    Improved Local Tangent Space Alignment

    do.isomap()

    Isometric Feature Mapping

    do.ispe()

    Isometric Stochastic Proximity Embedding

    do.keca()

    Kernel Entropy Component Analysis

    do.klde()

    Kernel Local Discriminant Embedding

    do.klfda()

    Kernel Local Fisher Discriminant Analysis

    do.klsda()

    Kernel Locality Sensitive Discriminant Analysis

    do.kmfa()

    Kernel Marginal Fisher Analysis

    do.kmmc()

    Kernel Maximum Margin Criterion

    do.kpca()

    Kernel Principal Component Analysis

    do.kqmi()

    Kernel Quadratic Mutual Information

    do.ksda()

    Kernel Semi-Supervised Discriminant Analysis

    do.lamp()

    Local Affine Multidimensional Projection

    do.lapeig()

    Laplacian Eigenmaps

    do.lisomap()

    Landmark Isometric Feature Mapping

    do.lle()

    Locally Linear Embedding

    do.llle()

    Local Linear Laplacian Eigenmaps

    do.ltsa()

    Local Tangent Space Alignment

    do.mmds()

    Metric Multidimensional Scaling

    do.mve()

    Minimum Volume Embedding

    do.mvu()

    Maximum Variance Unfolding / Semidefinite Embedding

    do.nnp()

    Nearest Neighbor Projection

    do.phate()

    Potential of Heat Diffusion for Affinity-based Transition Embedding

    do.plp()

    Piecewise Laplacian-based Projection (PLP)

    do.ree()

    Robust Euclidean Embedding

    do.rpca()

    Robust Principal Component Analysis

    do.sammon()

    Sammon Mapping

    do.sne()

    Stochastic Neighbor Embedding

    do.spe()

    Stochastic Proximity Embedding

    do.splapeig()

    Supervised Laplacian Eigenmaps

    do.spmds()

    Spectral Multidimensional Scaling

    do.tsne()

    t-distributed Stochastic Neighbor Embedding

  • Out-of-sample Prediction
  • oos.linproj()

    OOS : Linear Projection

  • Auxiliary Functions
  • aux.gensamples()

    Generate model-based samples

    aux.graphnbd()

    Construct Nearest-Neighborhood Graph

    aux.kernelcov()

    Build a centered kernel matrix K

    aux.pkgstat()

    Show the number of functions for Rdimtools.

    aux.preprocess()

    Preprocessing the data

    aux.shortestpath()

    Find shortest path using Floyd-Warshall algorithm

  • Data
  • iris

    Load Iris data

    usps

    Load USPS handwritten digits data