contents.m

偏最小二乘算法在MATLAB中的实现

% PLS_Toolbox.
% Version 1.5.1  22-October-95
% Copyright (c) 1995 by Eigenvector Technologies
% Barry M. Wise and Neal B. Gallagher
%
% Data Scaling and Preprocessing.
%   auto     - Autoscales data.
%   mncn     - Mean centers data.
%   scale    - Scales data using specified means and std. devs.
%   rescale  - Scales data back to original scaling.
%   delsamps - Deletes samples from data matrices.
%   savgol   - Savitsky-Golay smoothing and derivatives.
%   shuffle  - Randomly re-orders matrix rows.
%   dp       - Adds a diagonal line to prediction plots.
%
% Elementary Statistics.
%   ttestp   - t test and inverse t test statistic.
%   ftest    - Inverse F test statistic.
%   anova1w  - One-way analysis of variance.
%   anova2w  - Two-way analysis of variance.
%
% Principal Components, Cluster and Evolving Factor Analysis.
%   pca      - Principal components analysis.
%   pcapro   - Applies existing PCA model to new data.
%   mdpca    - PCA for matrices with missing data.
%   bigpca   - PCA for large matrices.
%   pltloads - Two and three dimensional loadings plots.
%   pltscrs  - Two and three dimensional scores plots.
%   cluster  - K-means and KNN cluster analysis with dendrograms.
%   gcluster - Graphical user interface for cluster
%   evolvfa  - Evolving factor analysis
%
% Principal Components Regression.
%   pcr1     - Principal components regression.
%   pcrcv    - General cross-validation for PCR models.
%   pcrcv1   - Leave one out cross-validation for PCR models.
%   pcrcvblk - PCR cross validation using contiguous data blocks.
%   pcrcvrnd - PCR cross validation using SDEP.
%
% Partial Least Squares Regression.
%   pls      - Partial least squares regression.
%   plsnipal - NIPALS algorithm for one PLS latent variable.
%   plspred  - Predictions based on existing PLS model.
%   conpred  - Converts PLS models to regression vectors.
%   conpred1 - Converts PLS models to single regression vector.
%   plscv    - General cross-validation for PLS models.
%   plscv1   - Leave one out cross-validation for PLS models.
%   pcrcvblk - PLS cross validation using contiguous data blocks.
%   pcrcvrnd - PLS cross validation using SDEP.
%   simpls1  - PLS by SIMPLS for univariate Y.
%
% Other Linear Regression Methods.
%   cr       - Continuum regression by SIMPLS algorithm
%   crcvrnd  - Cross-validation for continuum regression.
%   powerpls - CR by "powered" method - obsolete, use cr.
%   ridge    - Ridge regression by Hoerl-Kennard method.
%   ridgecv  - Ridge regression by cross-validation.
%
% General Regression Modelling.
%   modlmker - Develops PCR, PLS and RR models.
%   modlrder - Displays information from MODLMKER models.
%   modlpred - Predictions based on models created by MODLMKER.
%   modlgui  - Graphical user interface for modlmker  
%
% Non-Linear Regression Methods.
%   polypls  - PLS with polynomial inner-relationships.
%   polypred - New predictions with poly-PLS models.
%   spl_pls  - PLS with spline inner-relationships.
%   splspred - New predictions with SPL_PLS.
%   splnfit  - Spline fits to bivariate data.
%   splnpred - New predictions based on spline fits.
%   lwrpred  - Predictions based on LWR model.
%   lwrxy    - LWR predictions with y-distance weighting.
%
% PLS with Neural Network Inner Relationships Functions.
%   nnpls    - Cross-validation of NN-PLS models.
%   collapse - Converts NN-PLS models to NN form.
%   nnplsbld - Parameterizes NN-PLS models given form.
%   nnplsprd - Predictions from collapsed NN-PLS models.
%
% Multivariate Instrument Standardization.
%   stdgen   - Instrument standardization transform generator.
%   stdsslct - Selects data subsets for use in standardization.
%
% Multivariate Statistical Process Control.
%   replace  - Replaces variables based on PCA or PLS models.
%   plsrsgn  - Generates a matrix of PLS models for MSPC.
%   plsrsgcv - Cross-validation for PLSRSGN models.
%
% Identification of Finite Impulse Response Models. 
%   plspuls  - Identifies FIR models by PLS for MISO systems.
%   fir2ss   - Transforms FIR model to equiv. state space model.
%   writein2 - Writes matrices for dynamic model identification.
%   wrtpulse - Writes matrices with delays for identification.
%   autocor  - Auto-correlation function for time series data.
%   crosscor - Cross-correlation function for time series data.
%
% PLS_Toolbox Demonstrations.
%   statdemo - t test, F test, one and two-way ANOVA.
%   pcademo  - Pricipal components analysis.
%   efa_demo - Evolving factor analysis.
%   mddemo   - PCA for missing data.
%   clstrdmo - Statistical cluster analysis and dendrograms.
%   plsdemo  - Partial least squares regression and PCR.
%   ridgdemo - Ridge regression functions.
%   pwrdemo  - Continuum regression by "power" method.
%   modldemo - PLS, PCR and RR modelling with MODLMKER.
%   polydemo - PLS with polynomial inner relationship.
%   lwrdemo  - Locally weighted regression functions.
%   splndemo - PLS with spline inner relationships.
%   stddemo  - Multivariate instrument standardization.
%   sgdemo   - Savitsky-Golay smoothing and derivatives.
%   ccordemo - Cross- and auto-correlation.
%   pulsdemo - Identification of FIR models with PLS.
%   rsgndemo - Collections of PLS models for MSPC.
%   rplcdemo - Replacement of failed sensors with MSPC models.
%
% PLS_Toolbox Test Data Sets.
%   splndata - Synthetic data for spline demo.
%   nir_data - Near infrared spectra of pseudo gasoline samples.
%   pcadata  - Liquid-fed ceramic melter data for pca demo.
%   ridgdata - Hald data set of cement samples.
%   repdata  - Liquid-fed ceramic melter data for replace demo.
%   pol_data - Non-linear surge tank data for polypls demo.
%   pulsdata - Liquid-fed ceramic melter data for plspulsm demo.
%   plsdata  - Liquid-fed ceramic melter data for pls demo.
%   statdata - Data sets for statdemo
%
% Further Information.
%   readme   - Release notes on PLS_Toolbox 1.5