regression

(To be removed) Perform linear regression of shallow network outputs on targets

regression will be removed. For more information, see Version History.

Syntax

[r,m,b] = regression(t,y)

[r,m,b] = regression(t,y,'one')

Description

[r,m,b] = regression(t,y) calculates the linear regression between each element of the network response and the corresponding target.

This function takes cell array or matrix target t and output y, each with total matrix rows of N, and returns the regression values, r, the slopes of regression fit, m, and the y-intercepts, b, for each of the N matrix rows.

[r,m,b] = regression(t,y,'one') combines all matrix rows before regressing, and returns single scalar regression, slope, and offset values.

Examples

Fit Regression Model and Plot Fitted Values versus Targets

This example shows how to train a feedforward network and calculate and plot the regression between its targets and outputs.

Load the training data.

[x,t] = simplefit_dataset;

The 1-by-94 matrix x contains the input values and the 1-by-94 matrix t contains the associated target output values.

Construct a feedforward neural network with one hidden layer of size 20.

net = feedforwardnet(20);

Train the network net using the training data.

net = train(net,x,t);

Estimate the targets using the trained network.

y = net(x);

Calculate and plot the regression between its targets and outputs.

[r,m,b] = regression(t,y)

r = 
1.0000

m = 
1.0000

b = 
1.0878e-04

plotregression(t,y)

Input Arguments

t — Target

matrix | cell array

Network targets, specified as a matrix or cell array.

y — Output

scalar

Network outputs, specified as a matrix or cell array.

Output Arguments

r — Regression value

scalar

Regression value, returned as a scalar.

m — Slope

scalar

Slope of regression fit, returned as a scalar.

b — Offset

scalar

Offset of regression fit, returned as a scalar.

Version History

Introduced in R2010b

R2026a: regression will be removed

regression will be removed in a future release.

To train a neural network with tabular data, use one of these functions:

• fitrnet (Statistics and Machine Learning Toolbox) — Train regression neural network. Use this function for simple regression tasks.

• fitcnet (Statistics and Machine Learning Toolbox) — Train classification neural network. Use this function for simple classification tasks.

• trainnet — Train neural network. Use this function when the fitrnet and fitcnet functions do not provide the training options that you need for your task.

To train a neural network with time series data, use one of these options:

• Time Series Modeler app — Interactively design and train neural networks for time series data.

• trainnet — Train neural network. Use this function when you need more flexibility in the network architecture or training algorithm.

To design and customize your own neural network for these workflows, you can create a network using an array of deep learning layers or a dlnetwork object. To create and edit neural networks interactively and generate code, use the Deep Network Designer app. To learn more about neural network building workflows, see these examples and topics:

• Build Networks with Deep Network Designer

• List of Deep Learning Layers

• Example Deep Learning Network Architectures

• Multiple-Input and Multiple-Output Networks

• Define Custom Deep Learning Layers

For more information, see Transition Legacy Neural Network Code to dlnetwork Workflows.

R2020b: regression is not recommended

regression is not recommended. To fit a linear regression model, use fitlm (Statistics and Machine Learning Toolbox) instead.