Class Summary |
AttrEvalMetricLearner |
AttrEvalMetricLearner - sets the weights of a metric
using scores from an attribute evaluator |
BarHillelMetric |
Class for performing RCA according to Bar-Hillel's algorithm. |
BarHillelMetricMatlab |
Class for performing RCA according to Bar-Hillel's algorithm. |
ClassifierMetricLearner |
ClassifierMetricLearner - learns metric parameters by constructing
"difference instances" and then learning weights that classify same-class
instances as positive, and different-class instances as negative. |
GDMetricLearner |
GDMetricLearner - sets the weights of a metric
using gradient descent |
HardPairwiseSelector |
HardPairwiseSelector class. |
KL |
KL class
Implements weighted Kullback-Leibler divergence |
LearnableMetric |
Interface to distance metrics that can be learned |
MatlabMetricLearner |
MatlabMetricLearner - learns metric parameters by constructing
"difference instances" and then learning weights that classify same-class
instances as positive, and different-class instances as negative using an
external Matlab program. |
Metric |
Abstract Metric class |
MetricLearner |
Abstract MetricLearner interface. |
PairwiseSelector |
Abstract PairwiseSelector class. |
RandomPairwiseSelector |
RandomPairwiseSelector class. |
TrainingPair |
This is a basic class for a distance training pair |
WeightedDotP |
WeightedDotP class
Implements the weighted dot product distance metric |
WeightedEuclidean |
WeightedEuclidean class
Implements weighted euclidean distance metric |
WeightedMahalanobis |
WeightedMahalanobis class
Implements a weighted Mahalanobis distance metric weighted by a full matrix of weights. |
XingMetric |
Class for performing RCA according to Bar-Hillel's algorithm. |