| Name | Description |
|---|---|
| AddedNeuronGeneStruct | Represents an added neuron. When a neuron is added to a neural network in NEAT an existing connection between two neurons is discarded and replaced with the new neuron and two new connections, one connection between the source neuron and the new neuron and another from the new neuron to the target neuron. This struct represents those three IDs. This struct exists to represent newly added structure in a history buffer of added structures. This allows us to re-use IDs where a mutation recreates a structure that has previously occurred through previous mutations on other genomes. |
| ConnectionGene | A gene that represents a single connection between neurons in NEAT. |
| ConnectionGeneListBuilder | Used for building a list of connection genes. Connection genes are added one by one to a list and a dictionary of added connection genes is maintained keyed on ConnectionEndpointsStruct to allow a caller to check if a connection with the same end points (and potentially a different innovation ID) already exists in the list. |
| ConnectionMutationInfo | Each instance of ConnectionMutationInfo describes a type of mutation and associated parameters. A list of ConnectionMutationInfo objects describe a connection weight mutation scheme - a set of a number of different types of mutation along with the probability of each type occurring when spawning offspring genomes asexually. |
| CorrelationItem | A single comparison item resulting from the comparison of two genomes. If the CorrelationItemType is Match then both connection gene properties will be non-null, otherwise one of them will be null and the other will hold a reference to a disjoint or excess connection gene. Note. We generally only compare connection genes when comparing genomes. Connection genes along with their innovation IDs actually represent the complete network topology (and of course the connection weights). |
| CorrelationResults | The results from comparing two NEAT genomes and correlating their connection genes. |
| CorrelationStatistics | Statistics resulting from the comparison of two NEAT genomes. |
| NeatGenomeParameters | Represents parameters specific to NEAT genomes. E.g. parameters that describe probabilities for the different types of mutation and the proportion of possible connections to instantiate between input and output neurons within the initial population. |
| NeatGenomeStats | Various statistics for NeatGenome. |
| NeatGenomeXmlIO | Static class for reading and writing NeatGenome(s) to and from XML. |
| NeuronGene | A gene that represents a single neuron in NEAT. |