LogGenDisMixFunction

java.lang.Object
- de.jstacs.algorithms.optimization.DifferentiableFunction
- - de.jstacs.classifiers.differentiableSequenceScoreBased.OptimizableFunction
  - - de.jstacs.classifiers.differentiableSequenceScoreBased.AbstractOptimizableFunction
    - - de.jstacs.classifiers.differentiableSequenceScoreBased.AbstractMultiThreadedOptimizableFunction
      - de.jstacs.classifiers.differentiableSequenceScoreBased.DiffSSBasedOptimizableFunction
        
        de.jstacs.classifiers.differentiableSequenceScoreBased.gendismix.LogGenDisMixFunction

All Implemented Interfaces:

Function, MultiThreadedFunction

Direct Known Subclasses:

OneDataSetLogGenDisMixFunction
```
public class LogGenDisMixFunction
extends DiffSSBasedOptimizableFunction
```
This class implements the the following function
$\[f(\underline{\lambda}|C,D,\underline{\alpha},\underline{\beta})$
The weights have to sum to 1. For special weights the optimization turns out to be well known
- if the weights are (0,1,0), one obtains maximum likelihood,
- if the weights are (0,0.5,0.5), one obtains maximum a posteriori,
- if the weights are (1,0,0), one obtains maximum conditional likelihood,
- if the weights are (0.5,0,0.5), one obtains maximum supervised posterior,
- if the $\beta_2$ =0, one obtains the generative-discriminative trade-off,
- if the $\beta_2$ =0.5, one obtains the penalized generative-discriminative trade-off.
Of course, there are also some very interesting cases with other weights.

It can be used to maximize the parameters.

This class enables the user to exploit all CPUs of the computer by using threads. The number of compute threads can be determined in the constructor.

It is very important for this class that the DifferentiableSequenceScore.clone() method works correctly, since each thread works on its own clones.
Author:

Jens Keilwagen

Nested Class Summary
- Nested classes/interfaces inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.OptimizableFunction
  OptimizableFunction.KindOfParameter

Field Summary

Fields
Modifier and Type	Field and Description
`protected double[]`	`beta` The mixture parameters of the GenDisMix
`protected double[][]`	`cllGrad` Array for the gradient of the conditional log-likelihood
`protected double[][]`	`helpArray` General temporary array
`protected double[][]`	`llGrad` Array for the gradient of the log-likelihood
`protected double[]`	`prGrad` Array for the gradient of the prior

Fields inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.DiffSSBasedOptimizableFunction
dList, iList, prior, score, shortcut

Fields inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.AbstractMultiThreadedOptimizableFunction
params, worker

Fields inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.AbstractOptimizableFunction
cl, clazz, data, freeParams, logClazz, norm, sum, weights

Constructor Summary

Constructors
Constructor and Description
`LogGenDisMixFunction(int threads, DifferentiableSequenceScore[] score, DataSet[] data, double[][] weights, LogPrior prior, double[] beta, boolean norm, boolean freeParams)` The constructor for creating an instance that can be used in an `Optimizer`.

Method Summary

All Methods Instance Methods Concrete Methods
Modifier and Type	Method and Description
`protected void`	`evaluateFunction(int index, int startClass, int startSeq, int endClass, int endSeq)` This method evaluates the function for a part of the data.
`protected void`	`evaluateGradientOfFunction(int index, int startClass, int startSeq, int endClass, int endSeq)` This method evaluates the gradient of the function for a part of the data.
`protected double`	`joinFunction()` This method joins the partial results that have been computed using `AbstractMultiThreadedOptimizableFunction.evaluateFunction(int, int, int, int, int)`.
`protected double[]`	`joinGradients()` This method joins the gradients of each part that have been computed using `AbstractMultiThreadedOptimizableFunction.evaluateGradientOfFunction(int, int, int, int, int)`.
`void`	`reset(DifferentiableSequenceScore[] funs)` This method allows to reset the internally used functions and the corresponding objects.

Methods inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.DiffSSBasedOptimizableFunction
addTermToClassParameter, getClassParams, getDimensionOfScope, getParameters, reset, setParams, setThreadIndependentParameters

Methods inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.AbstractMultiThreadedOptimizableFunction
evaluateFunction, evaluateGradientOfFunction, getNumberOfAvailableProcessors, getNumberOfThreads, prepareThreads, setDataAndWeights, setParams, stopThreads

Methods inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.AbstractOptimizableFunction
getData, getParameters, getSequenceWeights

Methods inherited from class de.jstacs.algorithms.optimization.DifferentiableFunction
findOneDimensionalMin

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

- Field Detail
  - helpArray
```
protected double[][] helpArray
```
    General temporary array
  - llGrad
```
protected double[][] llGrad
```
    Array for the gradient of the log-likelihood
  - cllGrad
```
protected double[][] cllGrad
```
    Array for the gradient of the conditional log-likelihood
  - beta
```
protected double[] beta
```
    The mixture parameters of the GenDisMix
  - prGrad
```
protected double[] prGrad
```
    Array for the gradient of the prior
- Constructor Detail
  - LogGenDisMixFunction
```
public LogGenDisMixFunction(int threads,
                            DifferentiableSequenceScore[] score,
                            DataSet[] data,
                            double[][] weights,
                            LogPrior prior,
                            double[] beta,
                            boolean norm,
                            boolean freeParams)
                     throws IllegalArgumentException
```
    The constructor for creating an instance that can be used in an Optimizer.
    
    Parameters:
    
    threads - the number of threads used for evaluating the function and determining the gradient of the function
    
    score - an array containing the DifferentiableSequenceScores that are used for determining the sequences scores; if the weight beta[LearningPrinciple.LIKELIHOOD_INDEX] is positive all elements of score have to be DifferentiableStatisticalModel
    
    data - the array of DataSets containing the data that is needed to evaluate the function
    
    weights - the weights for each Sequence in each DataSet of data
    
    prior - the prior that is used for learning the parameters
    
    beta - the beta-weights for the three terms of the learning principle
    
    norm - the switch for using the normalization (division by the number of sequences)
    
    freeParams - the switch for using only the free parameters
    
    Throws:
    
    IllegalArgumentException - if the number of threads is not positive, the number of classes or the dimension of the weights is not correct
- Method Detail
  - joinGradients
```
protected double[] joinGradients()
                          throws EvaluationException
```
    Description copied from class: AbstractMultiThreadedOptimizableFunction
    
    This method joins the gradients of each part that have been computed using AbstractMultiThreadedOptimizableFunction.evaluateGradientOfFunction(int, int, int, int, int).
    
    Specified by:
    
    joinGradients in class AbstractMultiThreadedOptimizableFunction
    
    Returns:
    
    the gradient
    
    Throws:
    
    EvaluationException - if the gradient could not be evaluated properly
  - evaluateGradientOfFunction
```
protected void evaluateGradientOfFunction(int index,
                                          int startClass,
                                          int startSeq,
                                          int endClass,
                                          int endSeq)
```
    Description copied from class: AbstractMultiThreadedOptimizableFunction
    
    This method evaluates the gradient of the function for a part of the data.
    
    Specified by:
    
    evaluateGradientOfFunction in class AbstractMultiThreadedOptimizableFunction
    
    Parameters:
    
    index - the index of the part
    
    startClass - the index of the start class
    
    startSeq - the index of the start sequence
    
    endClass - the index of the end class (inclusive)
    
    endSeq - the index of the end sequence (exclusive)
  - joinFunction
```
protected double joinFunction()
                       throws DimensionException,
                              EvaluationException
```
    Description copied from class: AbstractMultiThreadedOptimizableFunction
    
    This method joins the partial results that have been computed using AbstractMultiThreadedOptimizableFunction.evaluateFunction(int, int, int, int, int).
    
    Specified by:
    
    joinFunction in class AbstractMultiThreadedOptimizableFunction
    
    Returns:
    
    the value of the function
    
    Throws:
    
    DimensionException - if the parameters could not be set
    
    EvaluationException - if the gradient could not be evaluated properly
  - evaluateFunction
```
protected void evaluateFunction(int index,
                                int startClass,
                                int startSeq,
                                int endClass,
                                int endSeq)
                         throws EvaluationException
```
    Description copied from class: AbstractMultiThreadedOptimizableFunction
    
    This method evaluates the function for a part of the data.
    
    Specified by:
    
    evaluateFunction in class AbstractMultiThreadedOptimizableFunction
    
    Parameters:
    
    index - the index of the part
    
    startClass - the index of the start class
    
    startSeq - the index of the start sequence
    
    endClass - the index of the end class (inclusive)
    
    endSeq - the index of the end sequence (exclusive)
    
    Throws:
    
    EvaluationException - if the gradient could not be evaluated properly
  - reset
```
public void reset(DifferentiableSequenceScore[] funs)
           throws Exception
```
    Description copied from class: DiffSSBasedOptimizableFunction
    
    This method allows to reset the internally used functions and the corresponding objects.
    
    Specified by:
    
    reset in class DiffSSBasedOptimizableFunction
    
    Parameters:
    
    funs - the new instances
    
    Throws:
    
    Exception - if something went wrong

Class LogGenDisMixFunction

Nested Class Summary

Nested classes/interfaces inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.OptimizableFunction

Field Summary

Fields inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.DiffSSBasedOptimizableFunction

Fields inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.AbstractMultiThreadedOptimizableFunction

Fields inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.AbstractOptimizableFunction

Constructor Summary

Method Summary

Methods inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.DiffSSBasedOptimizableFunction

Methods inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.AbstractMultiThreadedOptimizableFunction

Methods inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.AbstractOptimizableFunction

Methods inherited from class de.jstacs.algorithms.optimization.DifferentiableFunction

Methods inherited from class java.lang.Object

Field Detail

helpArray

llGrad

cllGrad

beta

prGrad

Constructor Detail

LogGenDisMixFunction

Method Detail

joinGradients

evaluateGradientOfFunction

joinFunction

evaluateFunction

reset