Skip to main content

Advanced Dataset Settings

Advanced settings control dimensionality reduction, reference-spectrum preprocessing, and sample selection. They apply when a draft is finalized.

Principal Component Analysis​

Set Perform PCA to Yes to create a PCA transform and apply it to samples during dataset generation. Set Number of latent variables to the number of components to retain; the current editor accepts 2 through 100.

PCA should be fitted from training samples only. Image pixels follow that rule, but reference-spectrum fitting currently also includes signatures assigned to Validation. This is a current limitation. When PCA and Validation reference spectra are combined, do not treat the validation metrics as independent of the fitted transform.

PCA increases finalization time. Choose the component count for the signal and model you intend to use rather than treating one value as a universal default. When PCA analysis results are available, the visualization uses the leading components even if the generated dataset retains more.

Reference spectra​

These controls apply to signatures added on the Reference spectra tab:

  • Reference spectra normalization: None, Area, Max, Length, or Sum
  • Reference spectra resampling type: Interpolation or Convolution
  • Reference spectra resampling method: for Interpolation, Linear, Nearest, Nearest Up, Zero, Slinear, Quadratic, or Cubic

Resampling aligns reference spectra to the dataset's target wavelengths. The appropriate type and method depend on the signature and sensor response; verify the resulting spectra for the intended scientific workflow.

Sampling​

The current editor offers two sampling methods:

  • Random selects samples without replacement.
  • Interleaved selects evenly spaced positions through each class's sample sequence.

Class split type defines the sample unit:

  • Pixel keeps individual labeled pixels as samples.
  • Label selects labels and averages the pixels in each selected label into one spectrum.

For Automatic data assignment, this setting also determines whether the training/validation split operates on pixels or whole labels.

An Automatic Pixel split creates disjoint pixel samples, but pixels from the same label and image can enter both Training and Validation. Spatially adjacent hyperspectral pixels are often correlated, so this is not an independent test of scene-level generalization. A Label split keeps each label region together, but Training and Validation can still share an image. To evaluate generalization across scenes, sites, sensors, dates, or acquisition conditions, use Manual assignment and hold out whole images or acquisitions.

If a class has only one eligible image sample or reference signature, the current Automatic split can place the same sample in both Training and Validation. Add independent samples instead of interpreting that validation metric as held out.

Optional per-class caps limit Training, Validation, and Test samples. Min threshold per pixel and Max threshold per pixel exclude values outside the configured bounds during generation.