Main Content

Wide Data via Lasso and Parallel Computing

This example shows how to use lasso along with cross validation to identify important predictors.

Load the sample data and display the description.

load spectra
Description =

  11×72 char array

    '== Spectral and octane data of gasoline ==                              '
    '                                                                        '
    'NIR spectra and octane numbers of 60 gasoline samples                   '
    '                                                                        '
    'NIR:     NIR spectra, measured in 2 nm intervals from 900 nm to 1700 nm '
    'octane:  octane numbers                                                 '
    'spectra: a dataset array containing variables for NIR and octane        '
    '                                                                        '
    'Reference:                                                              '
    'Kalivas, John H., "Two Data Sets of Near Infrared Spectra," Chemometrics'
    'and Intelligent Laboratory Systems, v.37 (1997) pp.255-259              '

Lasso and elastic net are especially well suited for wide data, that is, data with more predictors than observations with lasso and elastic net. There are redundant predictors in this type of data. You can use lasso along with cross validation to identify important predictors.

Compute the default lasso fit.

[b fitinfo] = lasso(NIR,octane);

Plot the number of predictors in the fitted lasso regularization as a function of Lambda , using a logarithmic x -axis.


It is difficult to tell which value of Lambda is appropriate. To determine a good value, try fitting with cross validation.

[b fitinfo] = lasso(NIR,octane,'CV',10);
Elapsed time is 7.353767 seconds.

Plot the result.


Display the suggested value of Lambda .

ans =


Display the Lambda with minimal MSE.

ans =


Examine the quality of the fit for the suggested value of Lambda .

lambdaindex = fitinfo.Index1SE;
mse = fitinfo.MSE(lambdaindex)
df = fitinfo.DF(lambdaindex)
mse =


df =


The fit uses just 11 of the 401 predictors and achieves a small cross-validated MSE.

Examine the plot of cross-validated MSE.

% Use a log scale for MSE to see small MSE values better

As Lambda increases (toward the left), MSE increases rapidly. The coefficients are reduced too much and they do not adequately fit the responses. As Lambda decreases, the models are larger (have more nonzero coefficients). The increasing MSE suggests that the models are overfitted.

The default set of Lambda values does not include values small enough to include all predictors. In this case, there does not appear to be a reason to look at smaller values. However, if you want smaller values than the default, use the LambdaRatio parameter, or supply a sequence of Lambda values using the Lambda parameter. For details, see the lasso reference page.

Cross validation can be slow. If you have a Parallel Computing Toolbox license, speed the computation of cross-validated lasso estimate using parallel computing. Start a parallel pool.

mypool = parpool()
Starting parallel pool (parpool) using the 'local' profile ...
connected to 6 workers.

mypool = 

 Pool with properties: 

            Connected: true
           NumWorkers: 6
              Cluster: local
        AttachedFiles: {}
    AutoAddClientPath: true
          IdleTimeout: 30 minutes (30 minutes remaining)
          SpmdEnabled: true

Set the parallel computing option and compute the lasso estimate.

opts = statset('UseParallel',true);
[b fitinfo] = lasso(NIR,octane,'CV',10,'Options',opts);
Elapsed time is 3.799009 seconds.

Computing in parallel using two workers is faster on this problem.

Stop parallel pool.

Parallel pool using the 'local' profile is shutting down.

See Also

| | | |

Related Topics