edu.stanford.nlp.classify
Class LogConditionalEqConstraintFunction
java.lang.Object
edu.stanford.nlp.optimization.AbstractCachingDiffFunction
edu.stanford.nlp.classify.LogConditionalEqConstraintFunction
- All Implemented Interfaces:
- DiffFunction, Function, HasInitial
public class LogConditionalEqConstraintFunction
- extends AbstractCachingDiffFunction
Maximizes the conditional likelihood with a given prior.
Constrains parameters for the same history to sum to 1
Adapted from LogConditionalObjectiveFunction
- Author:
- Kristina Toutanova
Constructor Summary |
LogConditionalEqConstraintFunction(int numFeatures,
int numClasses,
int[][] data,
int[] labels)
|
LogConditionalEqConstraintFunction(int numFeatures,
int numClasses,
int[][] data,
int[] labels,
double sigma)
|
LogConditionalEqConstraintFunction(int numFeatures,
int numClasses,
int[][] data,
int[] labels,
int prior,
double sigma,
double epsilon)
|
Method Summary |
protected void |
calculate(double[] x1)
Calculate the value at x and the derivative and save them in the respective fields |
protected Index<IntTuple> |
createIndex()
create an index for each parameter - the prior probs and the features with all of their values |
int |
domainDimension()
Returns the number of dimensions in the function's domain |
protected int |
indexOf(int c)
|
protected int |
indexOf(int f,
int c,
int val)
|
double[] |
initial()
use a random starting point uniform -1 1 |
double[] |
priors(double[] x1)
|
double[][][] |
to3D(double[] x1)
|
Methods inherited from class java.lang.Object |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
NO_PRIOR
public static final int NO_PRIOR
- See Also:
- Constant Field Values
QUADRATIC_PRIOR
public static final int QUADRATIC_PRIOR
- See Also:
- Constant Field Values
HUBER_PRIOR
public static final int HUBER_PRIOR
- See Also:
- Constant Field Values
QUARTIC_PRIOR
public static final int QUARTIC_PRIOR
- See Also:
- Constant Field Values
numFeatures
protected int numFeatures
numClasses
protected int numClasses
data
protected int[][] data
labels
protected int[] labels
numValues
protected int[] numValues
LogConditionalEqConstraintFunction
public LogConditionalEqConstraintFunction(int numFeatures,
int numClasses,
int[][] data,
int[] labels)
LogConditionalEqConstraintFunction
public LogConditionalEqConstraintFunction(int numFeatures,
int numClasses,
int[][] data,
int[] labels,
double sigma)
LogConditionalEqConstraintFunction
public LogConditionalEqConstraintFunction(int numFeatures,
int numClasses,
int[][] data,
int[] labels,
int prior,
double sigma,
double epsilon)
domainDimension
public int domainDimension()
- Description copied from interface:
Function
- Returns the number of dimensions in the function's domain
- Specified by:
domainDimension
in interface Function
- Specified by:
domainDimension
in class AbstractCachingDiffFunction
- Returns:
- the number of domain dimensions
indexOf
protected int indexOf(int c)
- Returns:
- the index of the prior for class c
indexOf
protected int indexOf(int f,
int c,
int val)
createIndex
protected Index<IntTuple> createIndex()
- create an index for each parameter - the prior probs and the features with all of their values
to3D
public double[][][] to3D(double[] x1)
priors
public double[] priors(double[] x1)
calculate
protected void calculate(double[] x1)
- Description copied from class:
AbstractCachingDiffFunction
- Calculate the value at x and the derivative and save them in the respective fields
- Specified by:
calculate
in class AbstractCachingDiffFunction
initial
public double[] initial()
- use a random starting point uniform -1 1
- Specified by:
initial
in interface HasInitial
- Overrides:
initial
in class AbstractCachingDiffFunction
- Returns:
- a domain point
Stanford NLP Group