Machine Learning Pipelines

Execute and deploy end-to-end machine learning workflows as pipelines

A pipeline combines and organizes multiple steps of a data processing workflow. For machine learning, these steps include data preparation, feature engineering, feature selection, modeling, and postprocessing. Executing a pipeline applies each step to the data as it passes through the pipeline. You can construct a pipeline for only data preprocessing or only feature engineering. Alternatively, you can create a machine learning pipeline with multiple steps for data preprocessing, feature engineering, classification or regression, and inference together.

Objects

expand all

Pipeline Object

LearningPipeline Machine learning pipeline (Since R2026a)

Data Preprocessing Components

`equalWidthBinnerComponent`	Pipeline component for grouping data into equal-width bins (Since R2026a)
`frequencyEncoderComponent`	Pipeline component for frequency encoding categorical variables (Since R2026a)
`kmeansEncoderComponent`	Pipeline component for feature extraction using k-means clustering (Since R2026a)
`normalizerComponent`	Pipeline component for normalizing data (Since R2026a)
`observationImputerComponent`	Pipeline component for imputing missing values (Since R2026a)
`observationRemoverComponent`	Pipeline component for removing observations (Since R2026a)
`oneHotEncoderComponent`	Pipeline component for encoding categorical data into one-hot vectors (Since R2026a)
`outlierImputerComponent`	Pipeline component for imputing outlier values (Since R2026a)
`outlierRemoverComponent`	Pipeline component for removing outlier values (Since R2026a)
`pcaComponent`	Pipeline component for principal component analysis (PCA) (Since R2026a)
`quantileBinnerComponent`	Pipeline component for binning data based on quantiles (Since R2026a)
`ricaComponent`	Pipeline component for feature extraction using reconstruction independent component analysis (RICA) (Since R2026a)
`sparseFilterComponent`	Pipeline component for feature extraction using sparse filtering (Since R2026a)

Feature Selection and Engineering Components

`featureSelectionClassificationANOVAComponent`	Pipeline component for performing feature selection using ANOVA algorithm (Since R2026a)
`featureSelectionClassificationChi2Component`	Pipeline component for performing feature selection using chi-square tests (Since R2026a)
`featureSelectionClassificationKruskalWallisComponent`	Pipeline component for performing feature selection using Kruskal-Wallis test (Since R2026a)
`featureSelectionClassificationMRMRComponent`	Pipeline component for performing MRMR feature selection in classification workflow (Since R2026a)
`featureSelectionClassificationNCAComponent`	Pipeline component for performing feature selection using neighborhood component analysis (NCA) for classification (Since R2026a)
`featureSelectionClassificationReliefFComponent`	Pipeline component for performing feature selection using ReliefF algorithm (Since R2026a)
`featureSelectionRegressionFTestComponent`	Pipeline component for performing feature selection using F-tests (Since R2026a)
`featureSelectionRegressionMRMRComponent`	Pipeline component for performing MRMR feature selection in regression workflow (Since R2026a)
`featureSelectionRegressionNCAComponent`	Pipeline component for performing feature selection using neighborhood component analysis (NCA) for regression (Since R2026a)
`featureSelectionRegressionReliefFComponent`	Pipeline component for performing feature selection using RReliefF algorithm (Since R2026a)
`variableSelectorComponent`	Pipeline component for manual variable selection (Since R2026a)

Supervised Learning Components

Classification Components

`classificationDiscriminantComponent`	Pipeline component for discriminant analysis classification (Since R2026a)
`classificationECOCComponent`	Pipeline component for multiclass classification using error-correcting output codes (ECOC) model (Since R2026a)
`classificationEnsembleComponent`	Pipeline component for ensemble classification (Since R2026a)
`classificationGAMComponent`	Pipeline component for binary classification using generalized additive model (GAM) (Since R2026a)
`classificationKernelComponent`	Pipeline component for classification using Gaussian kernel with random feature expansion (Since R2026a)
`classificationKNNComponent`	Pipeline component for classification using k-nearest neighbor model (Since R2026a)
`classificationLinearComponent`	Pipeline component for binary classification of high-dimensional data using linear model (Since R2026a)
`classificationNaiveBayesComponent`	Pipeline component for multiclass classification using naive Bayes model (Since R2026a)
`classificationNeuralNetworkComponent`	Pipeline component for classification using neural network model (Since R2026a)
`classificationSVMComponent`	Pipeline component for one-class and binary classification using SVM classifier (Since R2026a)
`classificationTreeComponent`	Pipeline component for multiclass classification using binary decision trees (Since R2026a)

Regression Components

`regressionEnsembleComponent`	Pipeline component for regression using ensemble of learners (Since R2026a)
`regressionGAMComponent`	Pipeline component for generalized additive model (GAM) for regression (Since R2026a)
`regressionGPComponent`	Pipeline component for Gaussian process regression (GPR) (Since R2026a)
`regressionLinearComponent`	Pipeline component for regression of high-dimensional data using a linear model (Since R2026a)
`regressionKernelComponent`	Pipeline component for regression using Gaussian kernel model (Since R2026a)
`regressionNeuralNetworkComponent`	Pipeline component for regression using neural network model (Since R2026a)
`regressionSVMComponent`	Pipeline component for regression using a support vector machine (SVM) model (Since R2026a)
`regressionTreeComponent`	Pipeline component for regression using binary decision trees (Since R2026a)

Custom Components

functionComponent Pipeline component for custom function (Since R2026a)

Functions

expand all

Creation and Modification

Automatic Connection

`series`	Connect components in series to create pipeline (Since R2026a)
`parallel`	Connect components or pipelines in parallel to create pipeline (Since R2026a)
`insert`	Insert component or pipeline into existing pipeline (Since R2026a)
`replace`	Replace existing pipeline component with new component (Since R2026a)

Manual Connection

`add`	Add new component or pipeline to existing pipeline (Since R2026a)
`remove`	Remove existing components or pipelines from pipeline (Since R2026a)
`connect`	Create connections between pipeline components (Since R2026a)
`disconnect`	Remove connections between ports in pipeline (Since R2026a)

Hierarchy

expand Expand subpipelines in pipeline (Since R2026a)

Execution

`learn`	Initialize and evaluate pipeline or component (Since R2026a)
`run`	Execute pipeline or component for inference after learning (Since R2026a)
`prune`	Remove unnecessary components and dependencies from pipeline (Since R2026a)
`reset`	Reset pipeline or component (Since R2026a)
`crossvalidate`	Cross-validate pipeline (Since R2026a)

Visualization and Description

`view`	View diagram of pipeline inputs, outputs, components, and connections (Since R2026a)
`describe`	Display summary of pipeline components (Since R2026a)

Deployment

package Create deployable archive or standalone application from pipeline (Since R2026a)

Topics

Machine Learning Pipeline Phases
Understand the learn and run pipeline phases for local and deployed execution.

Featured Examples

Create Simple Classification Pipeline

Create, learn, and run a machine learning pipeline for SVM classification.

Open Live Script

Tune Pipeline Hyperparameters Using Cross-Validation

Use cross-validation to select a pipeline parameter value.

Open Live Script