kaz-Anova/StackNet

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1326 stars · 336 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

StackNet is a Java-based meta-modelling framework that implements Wolpert's stacked generalization across multiple layers, resembling a feedforward neural network built one layer at a time with any supervised learning algorithm. It can be run from the command line or embedded in Java code, and leverages H2O's algorithms.

Use cases

• stack multiple machine learning models for better accuracy
• build multi-level stacked ensembles for classification
• build stacked ensembles for regression problems
• run model stacking from the command line
• combine classifiers and regressors in a layered meta-model

When to choose

• you want to squeeze maximum predictive accuracy via stacked generalization
• you work on the JVM and want a ready-made stacking framework
• you prefer command-line model training without writing much code

When to avoid

• you need actively maintained tooling with recent updates
• you want a pure Python scikit-learn workflow (use pystacknet or scikit-learn instead)
• you need deep learning with backpropagation rather than layer-wise stacking

Facets

library · maturity maintenance

machine-learning cli machine-learning data-science jvm cli cross-platform stacked-generalization stacking ensemble-learning meta-modelling java h2o

Member repositories

For agents

markdown · JSON · MCP: product_card(name="kaz-Anova/StackNet")