Search Algorithms |
This topic contains the following sections:
SearchAlgorithm includes GridSearch and RandomSearch implementations. These are very simple algorithms can be used for other optimization algorithms benchmarking and simple optimization problems solving.
The GridSearch algorithm consistently searches for the best objective value over finite size grid. So, this algorithm supports only SearchAlgorithmICountableFactor parameters (SearchAlgorithmCategoricalFactor and SearchAlgorithmDiscreteFactor classes). The parameters combination occurs only once on the all optimization run.
The RandomSearch algorithm randomly searches for the best objective value. Comparing to the GridSearch this algorithm supports also SearchAlgorithmContinuousFactor type. The disadvantage of this class is duplicate parameters combination possibility.
Next most important methods and properties are featured in the class:
Method | Description |
---|---|
The algorithm factors property. | |
Runs the search algorithm. | |
The minimal objective value have found. | |
The parameters point at which the minimal objective value have were found. | |
The exit conditions delegate. |