Package gov.sandia.cognition.learning.algorithm.confidence

Provides confidence-weighted categorization algorithms.

See: Description

Class Summary

Class	Description
AdaptiveRegularizationOfWeights	An implementation of the Adaptive Regularization of Weights (AROW) algorithm for online learning of a linear binary categorizer.
ConfidenceWeightedDiagonalDeviation	An implementation of the Standard Deviation (Stdev) algorithm for learning a confidence-weighted categorizer.
ConfidenceWeightedDiagonalDeviationProject	An implementation of the Standard Deviation (Stdev) algorithm for learning a confidence-weighted categorizer.
ConfidenceWeightedDiagonalVariance	An implementation of the Variance algorithm for learning a confidence-weighted linear categorizer.
ConfidenceWeightedDiagonalVarianceProject	An implementation of the Variance algorithm for learning a confidence-weighted linear categorizer.

Package gov.sandia.cognition.learning.algorithm.confidence Description

Provides confidence-weighted categorization algorithms. These methods are associate uncertainty with the estimated parameters, similar to Bayesian techniques. However, these techniques are not explicitly Bayesian in their formulation. In our test batteries, these methods tend to outperform techniques that do not propagate parameter uncertainty in their estimates.

Since:

3.0

Author:

Justin Basilico