Skip to content

Latest commit

 

History

History
119 lines (73 loc) · 5.57 KB

File metadata and controls

119 lines (73 loc) · 5.57 KB
graph LR
    neat_genome["neat.genome"]
    neat_genes["neat.genes"]
    neat_nn_feed_forward["neat.nn.feed_forward"]
    neat_nn_recurrent["neat.nn.recurrent"]
    neat_ctrnn["neat.ctrnn"]
    neat_iznn["neat.iznn"]
    neat_activations["neat.activations"]
    neat_aggregations["neat.aggregations"]
    neat_genome -- "creates and manipulates" --> neat_genes
    neat_genes -- "serve as building blocks for" --> neat_genome
    neat_genome -- "provides the blueprint for" --> neat_nn_feed_forward
    neat_nn_feed_forward -- "is configured by" --> neat_genome
    neat_genome -- "provides the blueprint for" --> neat_nn_recurrent
    neat_nn_recurrent -- "is configured by" --> neat_genome
    neat_genome -- "provides the blueprint for" --> neat_ctrnn
    neat_ctrnn -- "is configured by" --> neat_genome
    neat_genome -- "provides the blueprint for" --> neat_iznn
    neat_iznn -- "is configured by" --> neat_genome
    neat_nn_feed_forward -- "retrieves activation functions from" --> neat_activations
    neat_activations -- "provides functions to" --> neat_nn_feed_forward
    neat_nn_feed_forward -- "retrieves aggregation functions from" --> neat_aggregations
    neat_aggregations -- "provides functions to" --> neat_nn_feed_forward
    neat_nn_recurrent -- "retrieves activation functions from" --> neat_activations
    neat_activations -- "provides functions to" --> neat_nn_recurrent
    neat_nn_recurrent -- "retrieves aggregation functions from" --> neat_aggregations
    neat_aggregations -- "provides functions to" --> neat_nn_recurrent
    neat_ctrnn -- "retrieves activation functions from" --> neat_activations
    neat_activations -- "provides functions to" --> neat_ctrnn
    neat_ctrnn -- "retrieves aggregation functions from" --> neat_aggregations
    neat_aggregations -- "provides functions to" --> neat_ctrnn
    neat_iznn -- "retrieves activation functions from" --> neat_activations
    neat_activations -- "provides functions to" --> neat_iznn
    neat_iznn -- "retrieves aggregation functions from" --> neat_aggregations
    neat_aggregations -- "provides functions to" --> neat_iznn
Loading

CodeBoardingDemoContact

Details

The Genome & Network Representation subsystem is a core part of the neat-python project, an ML Toolkit/Library for Neuroevolution. This subsystem is responsible for defining the genetic encoding of neural networks (nodes and connections) and constructing executable neural network models from these genetic blueprints.

neat.genome

The central component representing the genetic blueprint of a neural network. It encapsulates the collection of node genes and connection genes, defining the network's topology and properties.

Related Classes/Methods:

neat.genes

Defines the fundamental genetic units (NodeGene, ConnectionGene) that compose a genome. These genes carry specific properties like activation functions, weights, and biases.

Related Classes/Methods:

neat.nn.feed_forward

Constructs and represents executable feed-forward neural networks based on a given neat.genome.

Related Classes/Methods:

neat.nn.recurrent

Constructs and represents executable recurrent neural networks from a neat.genome.

Related Classes/Methods:

neat.ctrnn

Implements and represents Continuous-Time Recurrent Neural Networks (CTRNNs), a specialized type of executable network model.

Related Classes/Methods:

neat.iznn

Implements and represents Izhikevich Spiking Neural Networks (IZNNs), another specialized executable network model.

Related Classes/Methods:

neat.activations

A registry and utility for managing and providing various activation functions (e.g., sigmoid, ReLU, tanh) used by the nodes within the neural networks.

Related Classes/Methods:

neat.aggregations

A registry and utility for managing and providing aggregation functions (e.g., sum, mean, max) used by the nodes to combine inputs from multiple connections.

Related Classes/Methods: