Version 4 (modified by mkofler, 14 years ago) (diff) |
---|
Analyzer
Analyzers are HeuristicLab 3.3 operators that implement the IAnalyzer interface. There are a couple of general purpose Analyzers, such as the QualityAnalyzer or the MultiAnalyzer as well as encoding and problem specific Analyzers.
1. General
1.1 MultiAnalyzer
An analyzer which applies arbitrary many other analyzers. It is used as a parameter in all currently implemented problems (Evolution Strategy?, Genetic Algorithm? etc.) to ensure easy extendability.
For example, if you open the GA for the ch130 problem? sample on the Optimizer Start Page, navigate to the Parameters tab and click on the Analyzer property, you will see the following list of sub-Analyzers:
The BestAverageWorstQualityAnalyzer comes from the genetic algorithm and the BestTSPSolutionAnalyzer was provided by the TSP problem. This wiring also works if you configure your own algorithm in the Optimizer. Opening a new Evolution Strategy algorithm and loading a Griewank test function as problem will result in a BestAverageWorstQualityAnalyzer and a BestSingleObjectiveTestFunctionSolutionAnalyzer. You can of course add additional Analyzers via the GUI.
Operator Parameters:
Parameter | Description |
---|---|
UpdateInterval | The interval in which the contained analyzers should be applied (Default: 1) |
UpdateCounter | The value which counts how many times the MultiAnalyzer was called since the last update. |
1.2 MinAverageMaxValueAnalyzer
An operator which analyzes the minimum, average and maximum of a value in the scope tree.
1.3 BestAverageWorstQualityAnalyzer
An operator which analyzes the best, average and worst quality of solutions in the scope tree.
1.4 QualityAnalyzer
An operator which analyzes the quality of solutions in the scope tree.
1.5 ValueAnalyzer
An operator which analyzes a value in the scope tree.
2. Tabu Search
2.1 TabuNeighborhoodAnalyzer
Operator Parameters:
Parameter | Description |
---|---|
IsTabu | A value that determines if a move is tabu or not. |
PercentTabu | Indicates how much of the neighborhood is tabu. |
Results | The result collection where the value should be stored. |
3. Artificial Ant Problem
3.1 BestAntTrailAnalyzer
An operator for analyzing the best ant trail of an Artificial Ant Problem?.
Operator Parameters:
Parameter | Description |
---|---|
BestSolution | The visual representation of the best ant trail. |
MaxTimeSteps | The maximal time steps that the artificial ant has available to collect all food items. |
Quality | The qualities of the artificial ant solutions which should be visualized. |
Results | The result collection where the best artificial ant solution should be stored. |
SymbolicExpressionTree | The artificial ant solutions from which the best solution should be visualized. |
World | The world with food items for the artificial ant. |
4. Knapsack Problem
4.1 BestKnapsackSolutionAnalyzer
An operator for analyzing the best solution for a Knapsack Problem?.
Operator Parameters:
Parameter | Description |
---|---|
BestKnownQuality | The quality of the best known solution. |
BestKnownSolution | The best known solution. |
BestSolution | The best knapsack solution. |
BinaryVector | The knapsack solutions from which the best solution should be visualized. |
KnapsackCapacity | Capacity of the Knapsack. |
Maximization | True if the problem is a maximization problem. |
Quality | The qualities of the knapsack solutions which should be visualized. |
Results | The result collection where the knapsack solution should be stored. |
Values | The values of the items. |
Weights | The weights of the items. |
5. OneMax Problem
5.1 BestOneMaxSolutionAnalyzer
An operator for analyzing the best solution for a OneMax Problem.
Operator Parameters:
Parameter | Description |
---|---|
BestKnownQuality | The quality of the best known solution. |
BestKnownSolution | The best known solution. |
BinaryVector | The Onemax solutions from which the best solution should be visualized. |
Maximization | True if the problem is a maximization problem. |
Quality | The qualities of the Onemax solutions which should be visualized. |
Results | The result collection where the Onemax solution should be stored. |
6. Single Objective Test Function Problem
6.1 BestSingleObjectiveTestFunctionSolutionAnalyzer
An operator for analyzing the best solution for a SingleObjectiveTestFunction problem.
7. Travelling Salesman Problem
7.1 BestTSPSolutionAnalyzer
An operator for analyzing the best solution of a Travelling Salesman Problem? given in path representation using city coordinates.
Attachments (1)
- MultiAnalyzer_Example_TSP.png (3.0 KB) - added by mkofler 14 years ago.
Download all attachments as: .zip