Struct SpikingNeuralClusterer 

pub struct SpikingNeuralClusterer { /* private fields */ }

Spiking neural network clusterer

This clusterer uses a network of spiking neurons with STDP learning to perform unsupervised clustering of spatial data. Input points are encoded as spike trains and presented to the network, which learns to respond selectively to different input patterns through competitive dynamics.

Features

• Rate coding for spatial data encoding
• STDP learning for adaptive weights
• Lateral inhibition for competitive dynamics
• Configurable network architecture
• Spike timing analysis

Example

use scirs2_core::ndarray::Array2;
use scirs2_spatial::neuromorphic::algorithms::SpikingNeuralClusterer;

let points = Array2::from_shape_vec((4, 2), vec![
    0.0, 0.0, 1.0, 0.0, 0.0, 1.0, 1.0, 1.0
]).unwrap();

let mut clusterer = SpikingNeuralClusterer::new(2)
    .with_spike_threshold(0.8)
    .with_stdp_learning(true)
    .with_lateral_inhibition(true);

let (assignments, spike_events) = clusterer.fit(&points.view()).unwrap();
println!("Cluster assignments: {:?}", assignments);

Implementations

impl SpikingNeuralClusterer

pub fn new(num_clusters: usize) -> Self

Create new spiking neural clusterer

Arguments

• num_clusters - Number of clusters to discover

Returns

A new SpikingNeuralClusterer with default parameters

pub fn with_spike_threshold(self, threshold: f64) -> Self

Configure spike threshold

Arguments

• threshold - Spike threshold for neurons

pub fn with_stdp_learning(self, enabled: bool) -> Self

Enable/disable STDP learning

Arguments

• enabled - Whether to enable STDP learning

pub fn with_lateral_inhibition(self, enabled: bool) -> Self

Enable/disable lateral inhibition

Arguments

• enabled - Whether to enable lateral inhibition

pub fn with_training_params( self, max_epochs: usize, simulation_duration: f64, ) -> Self

Configure training parameters

Arguments

• max_epochs - Maximum number of training epochs
• simulation_duration - Duration to simulate per data point

pub fn with_time_step(self, dt: f64) -> Self

Configure simulation time step

Arguments

• dt - Time step for simulation

pub fn fit( &mut self, points: &ArrayView2<'_, f64>, ) -> SpatialResult<(Array1<usize>, Vec<SpikeEvent>)>

Fit clustering to spatial data

Trains the spiking neural network on the provided spatial data using STDP learning and competitive dynamics to discover cluster structure.

Arguments

• points - Input points to cluster (n_points × n_dims)

Returns

Tuple of (cluster assignments, spike events) where assignments maps each point to its cluster and spike_events contains the complete spike timing history.

pub fn num_clusters(&self) -> usize

Get number of clusters

pub fn spike_threshold(&self) -> f64

Get spike threshold

pub fn is_stdp_enabled(&self) -> bool

Check if STDP learning is enabled

pub fn is_lateral_inhibition_enabled(&self) -> bool

Check if lateral inhibition is enabled

pub fn spike_history(&self) -> &[SpikeEvent]

Get current spike history

pub fn network_stats(&self) -> NetworkStats

Get network statistics

pub fn reset(&mut self)

Reset the network to initial state

Trait Implementations

impl Clone for SpikingNeuralClusterer

fn clone(&self) -> SpikingNeuralClusterer

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for SpikingNeuralClusterer

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.