Neural Networks

Neural networks for regression

Neural network models are structured as a series of layers that reflect the way the brain processes information. The regression neural network models available in Statistics and Machine Learning Toolbox™ are fully connected, feedforward neural networks for which you can adjust the size of the fully connected layers and change the activation functions of the layers.

To train a regression neural network model, use the Regression Learner app. For greater flexibility, train a regression neural network model using fitrnet in the command-line interface. After training, you can predict responses for new data by passing the model and the new predictor data to predict.

If you want to create more complex deep learning networks and have Deep Learning Toolbox™, you can try the Deep Network Designer (Deep Learning Toolbox) app.

Apps

Regression LearnerTrain regression models to predict data using supervised machine learning

Blocks

RegressionNeuralNetwork PredictPredict responses using neural network regression model (Since R2021b)

Functions

expand all

fitrnetTrain neural network regression model (Since R2021a)
compactReduce size of machine learning model
fitrqnetTrain regression quantile neural network (Since R2024b)
limeLocal interpretable model-agnostic explanations (LIME)
partialDependenceCompute partial dependence
permutationImportancePredictor importance by permutation (Since R2024a)
plotPartialDependenceCreate partial dependence plot (PDP) and individual conditional expectation (ICE) plots
shapleyShapley values (Since R2021a)
crossvalCross-validate machine learning model
kfoldLossLoss for cross-validated partitioned regression model
kfoldPredictPredict responses for observations in cross-validated regression model
kfoldfunCross-validate function for regression
kfoldLossLoss for cross-validated partitioned quantile regression model (Since R2025a)
kfoldPredictPredict responses for observations in cross-validated quantile regression model (Since R2025a)
kfoldfunCross-validate function for quantile regression (Since R2025a)
lossLoss for regression neural network (Since R2021a)
resubLossResubstitution regression loss
lossLoss for quantile neural network regression model (Since R2024b)
predictPredict responses using regression neural network (Since R2021a)
resubPredictPredict responses for training data using trained regression model
predictPredict response for quantile neural network regression model (Since R2024b)

Objects

RegressionNeuralNetworkNeural network model for regression (Since R2021a)
CompactRegressionNeuralNetworkCompact neural network model for regression (Since R2021a)
RegressionPartitionedNeuralNetworkCross-validated regression neural network model (Since R2023b)
RegressionQuantileNeuralNetworkQuantile neural network model for regression (Since R2024b)
CompactRegressionQuantileNeuralNetworkCompact quantile neural network model for regression (Since R2025a)
RegressionPartitionedQuantileModelCross-validated quantile model for regression (Since R2025a)

Topics