prtools.m

模式识别工具箱。非常丰富的底层函数和常见的统计识别工具

%parzenml    Optimization of smoothing parameter in Parzen density estimation.
%pca         Principal Component Analysis
%pcaklm      Back en routine for PC and KL mappings
%proxm       Proximity mapping and kernel construction
%reducm      Reduce to minimal space mapping
%remoutl     Remove outliers
%rejectm     Creates rejecting mapping
%scalem      Compute scaling data
%sigm        Simoid mapping
%spatm       Augment image dataset with spatial label information
%userkernel  User supplied kernel definition
%
%gtm         Fit a Generative Topographic Mapping (GTM) by EM
%plotgtm     Plot a Generative Topographic Mapping in 2D
%som         Simple routine computing a Self-Organizing Map (SOM)
%plotsom     Plot a Self-Organizing Map in 2D
%
%Classifier combiners
%--------------------
%averagec    Combining linear classifiers by averaging coefficients
%baggingc    Bootstrapping and aggregation of classifiers
%rsscc       Random subspace combining classifier
%votec       Voting classifier combiner
%wvotec      Weighted voting classifier combiner
%maxc        Maximum classifier combiner
%minc        Minimum classifier combiner
%meanc       Mean classifier combiner
%medianc     Median classifier combiner
%prodc       Product classifier combiner
%traincc     Train combining classifier
%fixedcc     Fixed combiner construction, back end
%parsc       Parse classifier or map
%rejectc     Creates reject version of exisiting classifier
%parallel    Parallel combining of classifiers
%stacked     Stacked combining of classifiers
%sequential  Sequential combining of classifiers
%
%
%Regression
%----------
%linearr     Linear regression
%ridger      Ridge regression
%lassor      LASSO
%svmr        Support vector regression
%ksmoothr    Kernel smoother
%knnr        k-nearest neighbor regression
%pinvr       Pseudo-inverse regression
%plsr        Partial least squares regression
%plsm        Partial least squares mapping
%
%testr       Mean squared regression error
%rsquared    R^2-statistic
%
%Handling images in datasets and datafiles
%-----------------------------------------
%data2im     Convert dataset to image
%getobjsize  Retrieve image size of feature images in datasets
%getfeatsize Retrieve image size of object images in datasets
%obj2feat    Transform object images to feature images in dataset
%feat2obj    Transform feature images to object images in dataset
%im2feat     Convert image to feature in dataset
%im2obj      Convert image to object in dataset
%selectim    Select image in multi-band object image dataset/datafile
%show        Display objects in datasets, datafiles and mappings
%
%Operations on images in datasets and datafiles
%----------------------------------------------
%classim     Classify image using a given classifier
%dataim      Image operation on dataset images (features or objects)
%doublem     Convert datafile images into double
%filtim      Image operation on objects in datafiles/datasets
%datfilt     Filter dataset image (outdated)
%datgauss    Filter dataset image by Gaussian filter
%datunif     Filter dataset image by uniform filter
%hsv2rgb     Convert HSV to RGB images
%rgb2hsv     Convert RGB to HSV images
%spatm       Augment image dataset with spatial label information
%im_bdilation      Binary dilation
%im_berosion       Binary erosion
%im_box            Bounding box
%im_bpropagation   Binary propagation
%im_center         Center image
%im_fft            FFT transform (and more)
%im_gaussf         Gaussian filtering
%im_gray           Multi-band to gray-value conversion
%im_hist_equalize  Histogram equalization
%im_invert         Invert image
%im_label          Labeling binary images
%im_maxf           Maximum filter
%im_minf           Minimum filter
%im_resize         Resize images
%im_rotate         Rotate images
%im_scale          Scale iamges
%im_select_blob    Select largest blob
%im_stretch        Contrast stretching of images
%im_threshold      Threshold images
%
%Feature extraction from images in datasets and datafiles
%--------------------------------------------------------
%histm       Convert images to histograms
%im_moments  Computes moments as features from object images
%im_mean     Computes center of gravity
%im_measure  Computes some measurements
%im_profile  Computes image profiles
%im_stat     Compute some simple statistics
%
%Clustering and distances
%------------------------
%distm       Distance matrix between two data sets.
%emclust     Expectation - maximization clustering
%proxm       Proximity mapping and kernel construction
%hclust      Hierarchical clustering
%kcentres    k-centres clustering
%kmeans      k-means clustering
%modeseek    Clustering by modeseeking
%
%mds         Non-linear mapping by multi-dimensional scaling (Sammon)
%mds_cs      Linear mapping by classical scaling
%mds_init    Initialisation of multi-dimensional scaling
%mds_stress  Dissimilarity of distance matrices
%
%Plotting
%--------
%gridsize    Set gridsize used in the PRTools plot commands
%plotc       Plot discriminant function for two features
%plote       Plot error curves
%plotf       Plot feature distribution
%plotm       Plot mapping
%ploto       Plot object functions
%plotr       Plot regression functions
%plotdg      Plot dendrgram (see hclust)
%scatterd    Scatterplot
%scatterdui  Scatterplot scatterplot with feature selection
%scatterr    Scatter regression dataset
%
%Various tests and support routines
%----------------------------------
%cdats       Support routine for checking datasets
%iscolumn    Test on column array
%iscomdset   Test on compatible datasets
%isdataim    Test on image dataset
%isdataset   Test on dataset
%isfeatim    Test on feature image dataset
%ismapping   Test on mapping
%isobjim     Test on object image dataset
%isparallel  Test on parallel mapping
%isstacked   Test on stacked mapping
%issym.m     Test on symmetric matrix
%isvaldset   Test on valid dataset
%isvaldfile  Test on valid datafile
%matchlablist Match entries of label lists
%newfig      Control of figures on the screen
%newline     Generate a new line in the command window
%labcmp      Compare two label lists and find the differences
%nlabcmp     Compare two label lists and count the differences
%testdatasize Check datasize and convert datafile to dataset
%
%Examples
%--------
%prex_cleval     learning curves
%prex_combining  classifier combining
%prex_confmat    confusion matrix, scatterplot and gridsize
%prex_datafile   datafile usage
%prex_datasets   standard datasets
%prex_density    Various density plots
%prex_eigenfaces Use of images and eigenfaces
%prex_matchlab   K-means clustering and matching labels
%prex_mcplot     Multi-class classifier plot
%prex_plotc      Dataset scatter and classifier plot
%prex_som        Training a SelfOrganizing Maps
%prex_spatm      Spatial smoothing of image classification
%prex_cost       Cost matrices and rejection
%prex_logdens    Density based classifier improvement
%prex_regr       Regression example
%
%prversion   returns version information on PRTools
%prprogress  report progress control
%prwarning   control PRTools warning level
%prmemory    controol PRTools large dataset handling
%prtver      prtools version back end
%typp        list prtools routine nicely

% Copyright: R.P.W. Duin, [e-mail]
% Faculty EWI, Delft University of Technology
% P.O. Box 5031, 2600 GA Delft, The Netherlands