Growing Neural Gas (GNG) on Iris Flowers

Growing Neural Gas (GNG) on Iris Flowers produced 10 clusters and 0 noise rows (silhouette 0.299, adjusted Rand 0.315 against held-out species).

Results of this run

Growing Neural Gas on Iris Flowers grows to maxNodes 10. Gas silhouette is 0.299, gas Calinski-Harabasz is 164.0, gas Davies-Bouldin is 0.989. Held-out species versus gas units: adjusted Rand 0.315. 0 noise rows out of 150 far from a unit. Gas unit 0 on Iris Flowers holds 8 rows (petal length (cm) 6.30 / sepal length (cm) 7.46). Gas unit 1 on Iris Flowers holds 13 rows (petal length (cm) 1.52 / sepal length (cm) 5.37). Gas unit 2 on Iris Flowers holds 14 rows (petal length (cm) 3.77 / sepal length (cm) 5.35). Gas unit 3 on Iris Flowers holds 17 rows (petal length (cm) 5.04 / sepal length (cm) 6.11). Gas unit 4 on Iris Flowers holds 16 rows (petal length (cm) 1.41 / sepal length (cm) 4.67). Gas unit 5 on Iris Flowers holds 18 rows (petal length (cm) 4.68 / sepal length (cm) 6.42). Gas unit 6 on Iris Flowers holds 21 rows (petal length (cm) 1.46 / sepal length (cm) 5.04). Gas unit 7 on Iris Flowers holds 11 rows (petal length (cm) 5.45 / sepal length (cm) 6.49). Gas unit 8 on Iris Flowers holds 18 rows (petal length (cm) 4.24 / sepal length (cm) 5.80). Gas unit 9 on Iris Flowers holds 14 rows (petal length (cm) 5.81 / sepal length (cm) 6.89). Node-budget path: max_nodes 6 grows 0.284 with 6 groups on Iris Flowers; max_nodes 10 grows 0.299 with 10 groups on Iris Flowers; max_nodes 16 grows 0.305 with 16 groups on Iris Flowers. Gas scatter keeps 95.8%.

Growing Neural Gas (GNG) on Iris Flowers: PCA plot of 10 clusters and 0 noise rows
Each point is one row. Gray points are noise. Computed 2026-11-02 with scikit-learn 1.4.2.
max_nodesSilhouetteClusters
60.2846
100.29910
160.30516
# Product Growing Neural Gas: max_nodes=10, random_state=42.
# Features are StandardScaler-transformed before the fit.

Who, how, and why

Method: StandardScaler, drop constant columns. Code version 3e4afc4b4352 (scripts/seo/build_explore_artifacts.py). scikit-learn 1.4.2. Seed 42. Dataset source: UCI Machine Learning Repository, Iris (Fisher, 1936). License: CC BY 4.0. The species column was held out of the fit and used only for the adjusted Rand score.

Results computed on 2026-11-02.

The results paragraph is filled from these numbers by a fixed template in the site code, so the scores can be checked against the same dataset in the tool.

Maintainer: nikolaou1982.

Contact: nick1982100@gmail.com.

How Growing Neural Gas (GNG) Works

Growing Neural Gas learns the topological structure of data by incrementally adding nodes and edges to a self-organizing network.

About the Iris Flowers Dataset

Classic 150-sample dataset with 4 petal/sepal measurements across 3 species. The gold standard for clustering & classification demos. Source: UCI Machine Learning Repository, Iris (Fisher, 1936). License: CC BY 4.0.

Samples. 150

Features. 4

Type. Numeric

Category. Probabilistic

When to Use Growing Neural Gas (GNG)

Growing Neural Gas (GNG) belongs to the Probabilistic family of clustering algorithms. These methods model each cluster as a probability distribution, providing soft assignments and uncertainty estimates.

Related Examples

Ready to try Growing Neural Gas (GNG) on your own data?

Upload a CSV and get instant results — free to start.