Overview   Package  Class  Use  Tree  Deprecated  Index  Help 
 PREV PACKAGE   NEXT PACKAGE FRAMES    NO FRAMES    

Package de.jstacs.algorithms.optimization.termination

Provides classes for termination conditions that can be used in algorithms.

See:
          Description

Interface Summary
TerminationCondition This interface can be used in any iterative algorithm for determining the end of the algorithm.
 

Class Summary
AbsoluteValueCondition Deprecated. use of the absolute value condition is not recommended and it may be removed in future releases
AbsoluteValueCondition.AbsoluteValueConditionParameterSet This class implements the parameter set for a AbsoluteValueCondition.
AbstractTerminationCondition This class is the abstract super class of many TerminationConditions.
AbstractTerminationCondition.AbstractTerminationConditionParameterSet This class implements the super class of all parameter sets of instances from AbstractTerminationCondition.
CombinedCondition This class allows to use many TerminationConditions at once.
CombinedCondition.CombinedConditionParameterSet This class implements the parameter set for a CombinedCondition.
IterationCondition This class will stop an optimization if the number of iteration reaches a given number.
IterationCondition.IterationConditionParameterSet This class implements the parameter set for a IterationCondition.
MultipleIterationsCondition This TerminationCondition requires another provided TerminationCondition to fail a contiguous specified number of times before the optimization is terminated.
MultipleIterationsCondition.MultipleIterationsConditionParameterSet This class implements the parameter set for a MultipleIterationsCondition.
SmallDifferenceOfFunctionEvaluationsCondition This class implements a TerminationCondition that stops an optimization if the difference of the current and the last function evaluations will be small, i.e., $|f(\underline{x}_{i-1}) - f(\underline{x}_i)| < \epsilon$.
SmallDifferenceOfFunctionEvaluationsCondition.SmallDifferenceOfFunctionEvaluationsConditionParameterSet This class implements the parameter set for a SmallDifferenceOfFunctionEvaluationsCondition.
SmallGradientConditon This class implements a TerminationCondition that allows no further iteration in an optimization if the the gradient becomes small, i.e., $\sum_i \left|\frac{\partial f(\underline{x})}{\partial x_i}\right| < \epsilon$.
SmallGradientConditon.SmallGradientConditonParameterSet This class implements the parameter set for a SmallStepCondition.
SmallStepCondition This class implements a TerminationCondition that allows no further iteration in an optimization if the scalar product of the current and the last values of x will be small, i.e., $(\underline{x}_i-\underline{x}_{i-1})^T (\underline{x}_i-\underline{x}_{i-1}) < \epsilon$.
SmallStepCondition.SmallStepConditionParameterSet This class implements the parameter set for a SmallStepCondition.
TimeCondition This class implements a TerminationCondition that stops the optimization if the elapsed time in seconds is greater than a given value.
TimeCondition.TimeConditionParameterSet This class implements the parameter set for a TimeCondition.
 

Package de.jstacs.algorithms.optimization.termination Description

Provides classes for termination conditions that can be used in algorithms.

See Also:
Optimizer, AbstractMixtureTrainSM, HigherOrderHMM
Overview   Package  Class  Use  Tree  Deprecated  Index  Help 
 PREV PACKAGE   NEXT PACKAGE FRAMES    NO FRAMES