MO-CMA-ES (Mutli-Objective Covariance Matrix Adaptation Evolution Strategy)
This is the implementation of the Mutli-Objective Covariance Matrix Adaptation Evolution Strategy, as published in Voss2010. The multi-objective covariance matrix adaptation evolution strategy (MO-CMA-ES) is an evolutionary algorithm for continuous vector-valued optimization. It combines indicator-based selection based on the contributing hypervolume with the efficient strategy parameter adaptation of the elitist covariance matrix adaptation evolution strategy (CMA-ES).
Usage
e["Solver"]["Type"] = "Optimizer/MOCMAES"
Results
These are the results produced by this solver:
- Best F(x)
Usage: e[“Results”][“Best F(x)”] = real number
Description: Optimal value of F(x) found so far.
- Best Parameters
Usage: e[“Results”][“Best Parameters”] = List of real number
Description: Value for the x parameters that produced the best F(x).
Variable-Specific Settings
These are settings required by this module that are added to each of the experiment’s variables when this module is selected.
- Lower Bound
Usage: e[“Variables”][index][“Lower Bound”] = real number
Description: [Hint] Lower bound for the variable’s value.
- Upper Bound
Usage: e[“Variables”][index][“Upper Bound”] = real number
Description: [Hint] Upper bound for the variable’s value.
- Initial Value
Usage: e[“Variables”][index][“Initial Value”] = real number
Description: [Hint] Initial value at or around which the algorithm shall start looking for an optimum.
- Initial Mean
Usage: e[“Variables”][index][“Initial Mean”] = real number
Description: [Hint] Initial mean for the proposal distribution. This value must be defined between the variable’s Mininum and Maximum settings (by default, this value is given by the center of the variable domain).
- Initial Standard Deviation
Usage: e[“Variables”][index][“Initial Standard Deviation”] = real number
Description: [Hint] Initial standard deviation of the proposal distribution for a variable (by default, this value is given by 30% of the variable domain width).
- Minimum Standard Deviation Update
Usage: e[“Variables”][index][“Minimum Standard Deviation Update”] = real number
Description: [Hint] Lower bound for the standard deviation updates of the proposal distribution for a variable. Korali increases the scaling factor sigma if this value is undershot.
- Values
Usage: e[“Variables”][index][“Values”] = List of real number
Description: [Hint] Locations to evaluate the Objective Function.
Configuration
These are settings required by this module.
- Population Size
Usage: e[“Solver”][“Population Size”] = unsigned integer
Description: Specifies the number of samples to evaluate per generation (preferably $4+3*log(N)$, where $N$ is the number of variables).
- Mu Value
Usage: e[“Solver”][“Mu Value”] = unsigned integer
Description: Number of best samples (offspring) advancing to the next generation (by default it is half the Sample Count).
- Evolution Path Adaption Strength
Usage: e[“Solver”][“Evolution Path Adaption Strength”] = real number
Description: Controls the learning rate of the conjugate evolution path (must be in (0,1], by default this variable is internally calibrated, variable Cc in reference).
- Covariance Learning Rate
Usage: e[“Solver”][“Covariance Learning Rate”] = real number
Description: Controls the learning rate of the covariance matrices (must be in (0,1], by default this variable is internally calibrated, variable Ccov in reference).
- Target Success Rate
Usage: e[“Solver”][“Target Success Rate”] = real number
Description: Value that controls the updates of the covariance matrix and the evolution path (must be in (0,1], variable Psucc in reference).
- Threshold Probability
Usage: e[“Solver”][“Threshold Probability”] = real number
Description: Threshold that defines update scheme for the covariance matrix and the evolution path (must be in (0,1], variable Pthresh in reference).
- Success Learning Rate
Usage: e[“Solver”][“Success Learning Rate”] = real number
Description: Learning Rate of success rates (must be in (0,1], by default this variable is internally calibrated, variable Cp in reference).
Termination Criteria
These are the customizable criteria that indicates whether the solver should continue or finish execution. Korali will stop when at least one of these conditions are met. The criteria is expressed in C++ since it is compiled and evaluated as seen here in the engine.
- Min Max Value Difference Threshold
Usage: e[“Solver”][“Min Max Value Difference Threshold”] = real number
Description: Specifies the min max fitness differential between two consecutive generations before stopping execution.
Criteria:
_k->_currentGeneration > 1 && (std::abs(*std::max_element(_currentBestValueDifferences.begin(), _currentBestValueDifferences.end())) < _minValueDifferenceThreshold)
- Min Variable Difference Threshold
Usage: e[“Solver”][“Min Variable Difference Threshold”] = real number
Description: Specifies the min L2 norm of the best samples between two consecutive generations before stopping execution.
Criteria:
_k->_currentGeneration > 1 && (*std::max_element(_currentBestVariableDifferences.begin(), _currentBestVariableDifferences.end()) < _minVariableDifferenceThreshold)
- Min Standard Deviation
Usage: e[“Solver”][“Min Standard Deviation”] = real number
Description: Specifies the minimal standard deviation.
Criteria:
_k->_currentGeneration > 1 && (*std::max_element(_currentMinStandardDeviations.begin(), _currentMinStandardDeviations.end()) <= _minStandardDeviation)
- Max Standard Deviation
Usage: e[“Solver”][“Max Standard Deviation”] = real number
Description: Specifies the maximal standard deviation.
Criteria:
_k->_currentGeneration > 1 && (*std::min_element(_currentMaxStandardDeviations.begin(), _currentMaxStandardDeviations.end()) >= _maxStandardDeviation)
- Max Value
Usage: e[“Solver”][“Max Value”] = real number
Description: Specifies the maximum target fitness to stop maximization.
Criteria:
_k->_currentGeneration > 1 && (+_bestEverValue > _maxValue)
- Min Value Difference Threshold
Usage: e[“Solver”][“Min Value Difference Threshold”] = real number
Description: Specifies the minimum fitness differential between two consecutive generations before stopping execution.
Criteria:
_k->_currentGeneration > 1 && (fabs(_currentBestValue - _previousBestValue) < _minValueDifferenceThreshold)
- Max Infeasible Resamplings
Usage: e[“Solver”][“Max Infeasible Resamplings”] = unsigned integer
Description: Maximum number of resamplings per candidate per generation if sample is outside of Lower and Upper Bound.
Criteria:
(_maxInfeasibleResamplings > 0) && (_infeasibleSampleCount >= _maxInfeasibleResamplings)
- Max Model Evaluations
Usage: e[“Solver”][“Max Model Evaluations”] = unsigned integer
Description: Specifies the maximum allowed evaluations of the computational model.
Criteria:
_maxModelEvaluations <= _modelEvaluationCount
- Max Generations
Usage: e[“Solver”][“Max Generations”] = unsigned integer
Description: Determines how many solver generations to run before stopping execution. Execution can be resumed at a later moment.
Criteria:
_k->_currentGeneration > _maxGenerations
Default Configuration
These following configuration will be assigned by default. Any settings defined by the user will override the given settings specified in these defaults.
{ "Covariance Learning Rate": -1.0, "Evolution Path Adaption Strength": -1.0, "Model Evaluation Count": 0, "Mu Value": 0, "Multinormal Generator": { "Mean Vector": [ 0.0 ], "Sigma": [ 1.0 ], "Type": "Multivariate/Normal" }, "Population Size": 0, "Success Learning Rate": 0.08, "Target Success Rate": 0.175, "Termination Criteria": { "Max Generations": 10000000000, "Max Infeasible Resamplings": 1000000, "Max Model Evaluations": 1000000000, "Max Standard Deviation": Infinity, "Max Value": Infinity, "Min Max Value Difference Threshold": -Infinity, "Min Standard Deviation": -Infinity, "Min Value Difference Threshold": -Infinity, "Min Variable Difference Threshold": -Infinity }, "Threshold Probability": 0.44, "Uniform Generator": { "Maximum": 1.0, "Minimum": 0.0, "Type": "Univariate/Uniform" }, "Variable Count": 0 }
Variable Defaults
These following configuration will be assigned to each of the experiment variables by default. Any settings defined by the user will override the given settings specified in these defaults.
{ "Initial Mean": NaN, "Initial Standard Deviation": NaN, "Initial Value": NaN, "Lower Bound": -Infinity, "Minimum Standard Deviation Update": 0.0, "Upper Bound": Infinity, "Values": [] }