Skip to content

Glossary

Term Meaning
ELM Extreme Learning Machine, a single hidden-layer network with random features and analytic output solve
OS-ELM Online Sequential ELM, an ELM updated chunk by chunk
ReOS-ELM Regularized OS-ELM with covariance regularization
FOS-ELM Forgetting OS-ELM with a forgetting factor
OS-CELM Constrained OS-ELM with class-distance constraint
ELM-AE ELM autoencoder used as a learned feature map
ML-ELM Multilayer ELM using stacked ELM-AE features and a batch ridge head
H-OS-ELM Hierarchical OS-ELM using stacked ELM-AE features and an online RLS head
FeatureMap Interface for transforming input matrices into feature matrices
Solver Strategy for learning output weights from features and targets
Backend CPU or GPU execution target
Ridge alpha Positive regularization added to least-squares normal systems
Forgetting factor RLS multiplier that down-weights stale online data
RBF Radial Basis Function feature using Euclidean distance to centers