Machine Learning Toolbox

Version 2.4, for MATLAB 9.9.0.1467703 (R2020b)

• • gabor2:
  • gabor_filter:
  • initfknn:
  • ldaOnWAndB:
  • myLinRegressEval: feature, each column is an observation
• Audio endpoint detection
  • epdPrmSet: Set parameters for endpoint detection
• Audio feature extraction
  • mfccOptSet: Set options for MFCC extraction from audio signals
  • wave2mfcc: MFCC (mel-scale frequency cepstral cofficient) extraction from audio signals
• Audio signal processing
  • enframe: Perform frame blocking on given signals
  • frame2sampleIndex: Frame index to sample index conversion
  • frameZeroMean: Make a frame zero mean by using polyfit
  • getTriFilterPrm: Get parameters of triangular filter bank
  • myAudioRead: Read an audio file to create an audio object
• Audio volume computation
  • frame2volume: Frame or frame matrix to volume conversion
• Classification analysis
  • confMatGet: Get confusion matrix from recognition results
  • confMatPlot: Display the confusion matrix
  • decisionBoundaryPlot: Plot of the decision boundary of a classifier
  • detDataGet: DET (Detection Error Tradeoff) data generation
  • detPlot: DET (Detection Error Tradeoff) plot for classification analysis of a single feature.
  • detPlotCore: Core function for plotting DET/ROC/PRC
  • prcDataGet: PRC (Precision-Recall Curve) data generation
  • prcPlot: ROC (Receiver Operating Characteristics) curve plot for classification analysis of a single feature.
  • rocPlot: ROC (Receiver Operating Characteristics) curve plot for classification analysis of a single feature.
• Classifier Evaluation
  • classifierEval: Evaluation of a given classifier
• Classifier plot
  • classifierPlot: Plot the results of a given classifier after training
• Classifier training
  • classifierTrain: Training a given classifier
• Coordinate transform
  • polarTransform: Polar transformation of an image
• Cost-sensitive classification
  • cost4binClassifier:
  • costVsTh4binClassifier:
• Data count reduction
  • dsCondense: Data condensing
  • dsEdit: Data editing
• Data dimension reduction
  • lda: Linear discriminant analysis
  • ldaFuzzy: fuzzy LDA (linear discriminant analysis)
  • ldaPerfViaKnncLoo: LDA recognition rate via KNNC and LOO performance index
  • pca: Principal component analysis (PCA)
  • pcaPerfViaKnncLoo: PCA analysis using KNNC and LOO
• Data fitting
  • oneCircleFit: Fit a circle via generalized least-squares method
• Dataset generation
  • dcData: Dataset generation for data clustering (no class label)
  • prData: Various test datasets for pattern recognition
• Dataset manipulation
  • dsClassMerge: Merge classes in a dataset
  • dsClassSize: Data count for each class for a data set
  • dsDataCheck: Check the data of the given dataset
  • dsDataClean: Check the format of the given dataset and do necessary cleaning
  • dsFeaDelete: Delete same-value features from a given dataset
  • dsNameAdd: Add names to a dataset if they are missing
  • inputNormalize: Input (feature) normalization to have zero mean and unity variance for each feature
  • inputPass: Pass the input to the output directly
  • inputWhiten: Whitening transformation based on eigen decomposition
• Dataset processing
  • dsDownSample: Down-sample a dataset
  • dsImpute: Dataset imputing for missing values
  • dsNanMap: Display NaN map for a given dataset
  • dsSubset: Subset of a dataset
  • dsSubset_old: Subset of a dataset
  • perfByField: Performance of classification based on values in an input field
• Dataset visualization
  • dsBoxPlot: Box plot for a dataset
  • dsCorrPlot: Box plot for a dataset
  • dsDistPlot: Plot the distribution of features in a data set
  • dsFeaVecPlot: Plot of feature vectors in each class
  • dsFeaVsIndexPlot: Plot of feature vs. data index
  • dsProjPlot1: Plot of output classes vs. each feature
  • dsProjPlot2: Plot of all possible 2D projection of the given dataset
  • dsProjPlot3: Plot of all possible 3D projection of the given dataset
  • dsRangePlot: Plot the range of features in a data set
  • dsScatterPlot: Scatter plot of the first 2 dimensions of the given dataset
  • dsScatterPlot3: Scatter plot of the first 3 dimensions of the given dataset
  • dsUniqueCount: Plot the count of unique values of each feature in a data set
• Distance and similarity
  • distLinScaling: Distance via linear scaling
  • distLpNorm: Lp norm of a vector
  • distLpNormPairwise: Pairwise Lp-norm distance between two set of vectors
  • distPairwise: Pairwise Euclidean distance between two set of vectors
  • distPairwiseM: Pairwise Euclidean distance between two set of vectors
  • distSqrPairwise: Pairwise squared Euclidean distance between two set of vectors
  • distSqrPairwiseM: Pairwise Euclidean distance between two set of vectors
  • simCos: Cosine of the angles betweeen two set of vectors
  • simSequence: Similarity between 2 sequences (e.g., for beat tracking)
• Dynamic time warping
  • dtw: DTW (dynamic time warping)
  • dtw1: DTW (dynamic time warping) with local paths of 27, 45, and 63 degrees
  • dtw1m: Pure m-file implementation of DTW (dynamic time warping) with local paths of 27, 45, and 63 degrees
  • dtw2: DTW (dynamic time warping) with local paths of 0, 45, and 90 degrees
  • dtw2m: Pure m-file implementation of DTW (dynamic time warping) with local paths of 0, 45, and 90 degrees
  • dtw3: DTW (dynamic time warping) with local paths of 0 and 45 degrees
  • dtw3m: Pure m-file implementation of DTW (dynamic time warping) with local paths of 0 and 45 degrees
  • dtw3withRestM: Pure m-file implementation of DTW (dynamic time warping) with local paths of 0 and 45 degrees
  • dtw4durationAlignment: Pure m-file implementation of DTW (dynamic time warping) for duration alignment
  • dtwBridgePlot: Bridge Plot of point-to-point mapping of DTW
  • dtwFixedPoint: Use of Picard iteration to find the optimal pitch shift for DTW
  • dtwOptSet: Set the parameters for DTW
  • dtwPathPlot: Plot the resultant path of DTW of two vectors
• Face recognition
  • faceData2ds: Convert face data to a dataset of the standard format
  • faceRecog: Face recognition via eigenfaces or fisherfaces
  • faceRecogDemo: Demo of face recognition using fisherfaces
  • faceRecogPerfLoo:
• Feature extraction
  • houghCircle:
• Fuzzy KNN
  • dsFuzzify: Initialize fuzzy membership grades from a crisp dataset.
• GMM
  • gmmEval: Evaluation of a GMM (Gaussian mixture model)
  • gmmGaussianNumEstimate: Estimate the best number of Gaussians via cross validation
  • gmmGrow: Increase no. of gaussian components within a GMM
  • gmmInitPrmSet: Set initial parameters for GMM
  • gmmRead: Read GMM parameters from a file
  • gmmTrain: GMM training for parameter identification
  • gmmWrite: Write the parameters of a GMM to a file
  • gmmcPlot: Plot the results of GMMC (Gaussian-mixture-model classifier)
• GMM classifier
  • gmmcEval: Evaluation of a GMM classifier with a given vector of priors
  • gmmcGaussianNumEstimate: GMM training and test, w.r.t. varying number of mixtures
  • gmmcTrain: Train a GMM classifier
• Gaussian PDF
  • gaussian: Multi-dimensional Gaussian propability density function
  • gaussianLog: Multi-dimensional log Gaussian propability density function
  • gaussianLogM: Multi-dimensional log Gaussian propability density function
  • gaussianM: Multi-dimensional Gaussian propability density function
  • gaussianMle: MLE (maximum likelihood estimator) for Gaussian PDF
  • gaussianSimilarity: Evaluation of a data set to see if it is close to a 1D Gaussian distribution
• HMM
  • dpOverMap: DP over matrix of state probability.
  • hmmEval: HMM evaluation
  • myPlateauPass: An m-file implementation of DP over matrix of state probability.
• Hierarchical clustering
  • hierClustering: Agglomerative hierarchical clustering
  • hierClusteringAnim: Display the cluster formation of agglomerative hierarchical clustering
  • hierClusteringPlot: Plot of the result from agglomerative hierarchical clustering, also known as dendrogram
• Image feature extraction
  • imFeaLbp: Local binary pattern for images
  • imFeaLgbp: Local Gabor binary pattern for images
• Input selection
  • inputSelectExhaustive: Input selection via exhaustive search
  • inputSelectExhaustive4reg: Input selection via exhaustive search for regression
  • inputSelectPlot: Plot for input selection
  • inputSelectSequential: Input selection via sequential forward selection
• Interpolation and regression
  • interpViaDistance: Interpolation via weighting of reciprocal distance
  • interpViaGaussian: Interpolation via normalized gaussian basis function
  • polyFitPiecewise: Piecewise polynomial fit (with 2 polynomials)
• K-nearest-neighbor classifier
  • classFuzzify: Initialize fuzzy membership grades for a dataset
  • knn: KNN (k-nearest-neighbor) search
  • knncEval: K-nearest neighbor classifier (KNNC)
  • knncFuzzy: Fuzzy k-nearest neighbor classifier
  • knncLoo: Leave-one-out recognition rate of KNNC
  • knncLooWrtK: Try various values of K in leave-one-out KNN classifier.
  • knncPlot: Plot the results of KNNC (k-nearest-neighbor classifier) after training
  • knncTrain: Training of KNNC (K-nearest neighbor classifier)
  • knncWrtK: Try various values of K in KNN classifier
• Least squares
  • planeFitViaTls: Plane fit via total lease-squares
• Least-squares estimate
  • lseTrainTest: Training & test procedure for LSE
  • perfLoo4lse: Leave-one-out cross validation for LSE (least-squares estimate) of A*x=b.
• Linear classifier
  • lincEval: Evaluation of linear classifier
  • lincOptSet: Set the training options for linear classifiers
  • lincTrain: Linear classifier (Perceptron) training
  • perceptronDemo: Demo of interactive perceptron training
  • quadraticDecBoundary:
• Multimedia data processing
  • dsCreateFromMm: Create DS from multimedia dataset
  • mmDataCollect: Collect multimedia data from a given directory
  • mmDataList: Collect multimedia data from a given directory
  • mmSetCollect: Collect multimedia data from a given directory
• NaN analysis
  • nanMapAnalysis: NaN analysis for a table
  • table2nanMap: NaN analysis for a table
• Naive Bayes classifier
  • nbcEval: Evaluation for the NBC (naive bayes classifier)
  • nbcPlot: Plot the results of NBC (naive Bayes classifier)
  • nbcTrain: Training the naive Bayes classifier (NBC)
• Nearest Neighbor Search
  • bbTreeGen: BB (branch-and-Bound) tree generation for nearest neighbor search
  • bbTreeSearch: BB (branch-and-bound) tree search for 1 nearest neighbor
• Object detection
  • beadDetect: Detect a single bead and return it's bounding circle
  • circleFit: Fitting a fixed number of circles in 2D via k-means-like clustering
  • circleFitValidate: Validate (determine) the number of circles for fitting in a 2D dataset
  • objDetect: Detect a single bead and return it's bounding circle
• Performance display
  • liftChart: Lift chart
• Performance evaluation
  • crossValidate: Cross validation for classifier performance evaluation
  • cvDataGen: Generate m-fold cross validation (CV) data for performance evaluation
  • perfCv: Cross-validation accuracy of given dataset and classifier
  • perfCv4classifier: Performance evaluation for various combinations of classifiers and input normalization schemes
  • perfLoo: Leave-one-out accuracy of given dataset and classifier
  • perfLoo4audio: Leave-one-file-out CV (for audio)
• Perfornamce analysis
  • perfAuc: DET (Detection Error Tradeoff) plot for classification analysis of a single feature.
• Quadratic classifier
  • lsecEval: Evaluation for the QC (quadratic classifier)
  • qcEval: Evaluation for the QC (quadratic classifier)
  • qcPlot: Plot the results of QC (quadratic classifier)
  • qcTrain: Training the quadratic classifier (QC)
• Sparse-representation classifier
  • srcEval: Evaluation of SRC (sparse-representation classifier)
  • srcTrain: Training the SRC (sparse-representation classifier)
• String match
  • dpPathPlot4strMatch: Plot the path of dynamic programming for string match.
  • editDistance: Edit distance (ED) via dynamic programming
  • lcs: Longest (maximum) common subsequence
  • xcorr4text: Cross correlation of two text strings
• String processing
  • strCentroid: Return the centroid of a set of strings
  • vecCentroid: Return the centroid of a set of strings
• Support vector machine
  • svmcEval: Evaluation of SVM (support vector machine) classifier
  • svmcTrain: Training SVM (support vector machine) classifier
• Utility
  • fileList: File list of given directories with a given extension name
  • getColor: Get a color from a palette
  • mixLogSum: Compute the mixture log sum
  • mltDoc: Online document of the given MLT command
  • mltRoot: Root of MLT (Machine Learning Toolbox)
  • mltSetup: Set up MLT toolbox
  • peaksFcn: Peaks function's value, gradient and Hessian.
  • toolboxInfo: Toolbox information
• Vector quantization
  • clusterValidate: Cluster validation
  • kMeansClustering: VQ (vector quantization) of K-means clustering using Forgy's batch-mode method
  • kMeansClusteringOnDist: K-means clustering on the distance matrix only
  • vecQuantize: Vector quantization using LBG method of center splitting
  • vqCenterInit: Find initial centers for VQ of k-means
  • vqCenterObjInit: Find the initial centers (objects) for K-means clustering with the distance matrix only
  • vqDataPlot: Plot the data and the result of vector quantization
• face classification
  • fcOptSet: Set FC (face classification) option
• least-square estimate classifier
  • lsecTrain: Training the LSE (least-square estimate) classifier (LSEC)