Package | Description |
---|---|
de.citec.tcs.alignment |
This module defines the interface for AlignmentAlgorithms as well as some helper classes.
|
de.citec.tcs.alignment.adp |
This package contains a more general approach to construct
AlignmentAlgorithms by relying on the theoretical concept of Algebraic
Dynamic Programming (ADP) as developed by Giegerich et al.
|
de.citec.tcs.alignment.comparators |
This module defines the interfaces for Comparators in the TCS Alignment Toolbox.
|
de.citec.tcs.alignment.learning |
This module is a custom implementation of the Large Margin
Nearest Neighbor classification scheme of Weinberger, Saul, et al.
|
de.citec.tcs.alignment.trees |
This package contains an implementation of the Tree Edit Distance according to Zhang & Shasha (1989).
|
Modifier and Type | Method and Description |
---|---|
double[] |
Alignment.computeGradient(DerivableComparator<X,Y> comp) |
double[] |
AlignmentList.computeGradient(DerivableComparator<X,Y> comp) |
double[] |
AlignmentMap.computeGradient(DerivableComparator<X,Y> comp) |
double[] |
CooptimalModel.computeGradient(DerivableComparator<X,Y> comp) |
double[] |
DerivableAlignmentDistance.computeGradient(DerivableComparator<X,Y> comp)
This computes the gradient of this DerivableAlignmentDistance w.r.t.
|
Constructor and Description |
---|
ParallelGradientEngine(Collection<Engine.CalculationResult<? extends MatrixEngine.MatrixCoordinate,? extends DerivableAlignmentDistance<X,Y>>> results,
int M,
int N,
DerivableComparator<X,Y> comparator)
Creates a ParallelGradientEngine that computes gradients for several
DeriableAlignmentDistance objects in parallel.
|
ParallelGradientEngine(DerivableAlignmentDistance<X,Y>[][] derivableMatrixEntries,
DerivableComparator<X,Y> comparator)
Creates a ParallelGradientEngine that computes gradients for several
DeriableAlignmentDistance objects in parallel.
|
ParallelGradientEngine(Map<MatrixEngine.MatrixCoordinate,DerivableAlignmentDistance<X,Y>> derivableMatrixEntries,
int M,
int N,
DerivableComparator<X,Y> comparator)
Creates a ParallelGradientEngine that computes gradients for several
DeriableAlignmentDistance objects in parallel.
|
Modifier and Type | Method and Description |
---|---|
double[] |
SoftADPPathModel.computeGradient(DerivableComparator<X,Y> comp) |
Modifier and Type | Class and Description |
---|---|
class |
AbstractDerivableValueComparatorWrapper<V>
An abstract convenience implementation of the ValueComparator interface and the
DerivableComparator interface.
|
class |
AlignmentSpecification
The alignment specification defines which keywords of the input sequences
shall be used and which comparator is to be used to calculate the local
distance between two values for that keyword.
|
class |
CharStatComparator
This compares strings using statistics over the number of letters inside.
|
class |
ComparisonBasedSkipExtendedComparator<X,Y>
This is an abstract class providing - in addition to the constant gap and skip costs provided by
the SkipExtendedComparator - the possibility to define gap and skip costs with respect to a
constant element from both input sets.
|
class |
DerivableStringComparatorWrapper
This is a wrapper for Comparators on StringValues.
|
class |
DerivableSymbolicComparatorWrapper
This is a wrapper for Comparators on SymbolicValues.
|
class |
DerivableVectorialComparatorWrapper
This is a wrapper for Comparators on VectorialValues.
|
class |
EuclideanComparator
Compares vectors using the L2-Norm.
|
class |
L1NormComparator
Compares vectors using the L1-Norm.
|
class |
NCDComparator
This compares strings using normalized compression distance.
|
class |
NormalizedEuclideanComparator
Implements a normalized L2 distance, defined as:
||x - y|| / (||x|| + ||y||) .
|
class |
NormalizedEuclideanRelevanceComparator
Implements a normalized relevance L2 distance, defined as:
||Λ * x - Λ * y|| / (||Λ * x|| + ||Λ * y||)
where Λ is a diagonal matrix.
|
class |
NormalizedL1Comparator
Implements a normalized L1 distance, defined as:
|x - y| / (|x| + |y|) .
|
class |
NormalizedL1RelevanceComparator
Implements a normalized relevance L1 distance, defined as:
|Λ * x - Λ * y| / (|Λ * x| + |Λ * y|)
where Λ is a diagonal matrix.
|
class |
ParameterLessComparisonBasedSkipExtendedComparator<X,Y>
A convenience implementation for ComparisonBasedSkipExtendedComparators without (own) parameters.
|
class |
ParameterLessSkipExtendedComparator<X,Y>
A convenience implementation for SkipExtendedComparators without (own) parameters.
|
class |
SkipExtendedComparator<X,Y>
This is an abstract class to be used for Comparators that are only to be used for comparisons per
se but can be extended for use with gaps and skips.
|
class |
SquareReplacementMatrixComparator<X>
This comparator defines explicit replacement (and deletion/insertion) costs for comparison of
values from a discrete set.
|
class |
TrivialEditComparator<X>
This is a Comparator for SymbolicValues that just specifies a score for:
the deletion of a symbol (1 per default)
the insertion of a symbol (1 per default)
the skip deletion of a symbol (0.5 per default)
the skip insertion of a symbol (0.5 per default)
a match between the same symbols (0 per default)
a mismatch between different symbols (1 per default)
Please note that this is just a special case of the much more general
ReplacementComparator.
|
class |
ZeroNormComparator
This comparator works on basis of the zero norm.
|
Modifier and Type | Method and Description |
---|---|
DerivableComparator<V,V> |
AbstractDerivableValueComparatorWrapper.getActualComparator() |
Constructor and Description |
---|
AbstractDerivableValueComparatorWrapper(String keyword,
ValueType type,
DerivableComparator<V,V> actualComparator) |
DerivableStringComparatorWrapper(String keyword,
DerivableComparator<String,String> actualComparator) |
DerivableSymbolicComparatorWrapper(String keyword,
DerivableComparator<String,String> actualComparator) |
DerivableVectorialComparatorWrapper(String keyword,
DerivableComparator<double[],double[]> actualComparator) |
Modifier and Type | Method and Description |
---|---|
double[] |
LMNNGradientCalculator.computeGradient(DerivableComparator<X,X> comp,
double[][] D)
Calculates the gradient of the LMNN cost function with respect to the
parameters of the given comparator.
|
Modifier and Type | Class and Description |
---|---|
class |
DerivableLabelComparator<X,Y>
This is a wrapper class which defined the distance between two tree nodes as the distance
between their labels.
|
Modifier and Type | Method and Description |
---|---|
double[] |
CooptimalMatrix.computeGradient(DerivableComparator<X,Y> comp) |
Constructor and Description |
---|
DerivableLabelComparator(DerivableComparator<X,Y> comparator) |
Copyright (C) 2016-2018 Benjamin Paaßen, AG Theoretical Computer Science, Centre of Excellence Cognitive Interaction Technology (CITEC), University of Bielefeld, licensed under the AGPL v. 3: http://openresearch.cit-ec.de/projects/tcs . This documentation is licensed under the conditions of CC-BY-SA 4.0: https://creativecommons.org/licenses/by-sa/4.0/