This task executes a Random Forest algorithm using multiple decision trees against the provided input raster, then performs classification. It creates a set of decision trees from a randomly selected subset of the training dataset.

This tool performs supervised classification on a single raster. You provide an input raster, ROIs, and parameter settings to generate a classified raster. For more advanced options, you can label data on one or more rasters in the Machine Learning Labeling Tool, train a model, and perform classification using the model in the Machine Learning Classification Tool, or build a workflow in the ENVI Modeler.

For background on the algorithm used, see Random Forest Classification.

  1. From the Toolbox, select Machine Learning > Supervised > Random Forest Classification. The Random Forest Classification dialog appears.
  2. Select an Input Raster, perform optional spatial and spectral subsetting and/or masking, then click OK.
  3. In the Input ROIs field, select an ROI file (.xml) that indicates the labeled pixels for the desired classes in the training raster. The ROIs must fall within the boundary of the input raster.
  4. Specify the ROI classes to use as background in the Background Labels field. These indicate classes of no interest.

  5. Select Yes or No for Balance Classes, to specify whether all classes should be considered equal during training. Selecting Yes helps to account for classes with few samples compared to classes with many samples.

  6. In the Max Features field, specify the number of features to consider when looking for the best split. The options are:

    • sqrt (default)

    • log2

  7. In the Custom Max Features field, specify the number of features to consider when looking for the best split. If specified, this value will override the Max Features setting.

  8. In the Estimators field, enter the number of decision trees to use. The estimators are the predictors of the algorithm. The default is 100.

  9. In the Max Depth field, specify the maximum depth of the tree. If not specified, then nodes are expanded until all leaves are pure.
  10. In the OOB field, specify whether to use out-of-bag samples to estimate the generalization score. The default is No.
  11. In the Output Raster field, enter a location and filename for the classification raster.In the Output Raster URI field, select a location and filename for the machine learning raster.
  12. Enable the Display result check box to display the output in the view when processing is complete.

  13. To reuse these task settings in future ENVI sessions, save them to a file. Click the down arrow and select Save Parameter Values, then specify the location and filename to save to. Note that some parameter types, such as rasters, vectors, and ROIs, will not be saved with the file. To apply the saved task settings, click the down arrow and select Restore Parameter Values, then select the file where you previously stored your settings.

  14. To run the process on a local or remote ENVI Server, click the down arrow and select Run Task in the Background or Run Task on remote ENVI Server name. The ENVI Server Job Console will show the progress of the job and will provide a link to display the result when processing is complete. See the ENVI Servers topic in ENVI Help for more information.

  15. To see a model-based version of this tool that shows how the tool is constructed from individual tasks, click Open in Modeler.

  16. Click OK.

See Also

ENVI Machine Learning Algorithms Background, TrainRandomForest Task, Extra Trees Classification Tool, K-Neighbors Classification Tool, Linear SVM Classification Tool, Naive Bayes Classification Tool, RBF SVM Classification Tool