Input
Executionexec_in
Execution trigger that begins Gaussian Mixture training
Fit Clustering (Gaussian Mixture) Node
AI/ML/Clustering
Fit/Train a Gaussian Mixture Model. Soft clustering with per-component covariances and mixture weights, fitted by Expectation-Maximization.
fit_gaussian_mixtureml
Ratings
Scores range from 0 to 10. Higher values mean more impact, exposure, or operational weight.
SecurityAttack surface and exposure impact.
6/10Medium
PrivacyPotential sensitivity of processed data.
6/10Medium
PerformanceRuntime or resource pressure.
4/10Medium
GovernancePolicy, audit, or compliance impact.
6/10Medium
ReliabilityOperational stability considerations.
5/10Medium
CostExternal or compute cost impact.
5/10Medium
Input Pins
10Data Source
Stringsource
Choose which backend supplies the training data
Default Database
Components
Integern_clusters
Number of Gaussian components (k) in the mixture. Each component costs a full d x d covariance matrix.
Default 3
Covariance Type
Stringcovariance_type
Shape of each component's covariance. linfa 0.8 implements full covariances only - scikit-learn's diag, tied and spherical variants do not exist here, so every component always costs d x d parameters.
Default Full
Init Method
Stringinit_method
How initial responsibilities are built: KMeans runs a KMeans pass first (usually the better optimum), Random draws them uniformly.
Default KMeans
Runs
Integern_runs
Number of EM passes. Note: linfa 0.8 continues each pass from the previous parameters instead of re-initializing, so this multiplies the iteration budget (Runs x Max Iterations) rather than performing independent restarts. Vary the Seed for a genuinely different start.
Default 1
Tolerance
Floattolerance
EM stops once the average log-likelihood gain per iteration falls below this value
Default 0.001
Reg Covariance
Floatreg_covariance
Non-negative value added to each covariance diagonal to keep it positive definite. Raise it when the fit reports a singular covariance; 0 makes duplicate or constant rows fail outright.
Default 0.000001
Max Iterations
Integermax_n_iterations
Maximum number of EM iterations per run
Default 100
Seed
Integerseed
Seed for the training row order. linfa 0.8 hard-codes its internal RNG (seed 42) and exposes no seeding hook on this entry point, so changing the seed re-orders the rows, which is what changes the initial responsibilities. Keep 42 to reproduce linfa's stock ordering.
Default 42
Output Pins
3Done
Executionexec_out
Activated once training completes
Model
Structmodel
Thread-safe handle to the trained Gaussian Mixture model
NodeMLModelNodeMLModel1 fields
model_refstringrequired
Mixture Weights
Float Arrayweights
Fitted mixture proportions, one per component, summing to 1. A tiny weight means that component captured almost no data.
Node Info
- Internal name
- fit_gaussian_mixture
- Category
- AI/ML/Clustering