| Class | Description |
|---|---|
| BestFirstStrategy |
This exploration strategy eventually explorers the whole design space but goes in the most promising directions
first, based on the
Fitness. |
| BreadthFirstStrategy |
A breadth-first search algorithm implementation, that
can work with multiple threads,
indeterministic,
saves all states (trajectories) as solutions that fulfill all the hard objectives,
can have a depth limit,
will backtrack when a model satisfies the hard objectives (after saving it as a solution) and will not explore
beyond that state.
|
| DepthFirstStrategy |
A depth-first search algorithm implementation, that
can work with multiple threads,
randomly traverses the search space,
saves all states (trajectories) as solutions that fulfill all the hard objectives,
can have a depth limit,
will backtrack when a model satisfies the hard objectives (after saving it as a solution), which can be modified
by calling
DepthFirstStrategy.continueIfHardObjectivesFulfilled()
|
| FixedPriorityStrategy |
Works as
DepthFirstStrategy but:
works only with single thread,
in a given state, it only traverses the activations with locally the highest priority.
|
| HillClimbingStrategy | |
| RandomSearchStrategy |