szilard/benchm-ml

A minimal benchmark for scalability, speed and accuracy of commonly used open source implementations (R packages, Python scikit-learn, H2O, xgboost, Spark MLlib etc.) of the top machine learning algorithms for binary classification (random forests, gradient boosted trees, deep neural networks etc.).

50
/ 100
Established

This project helps data professionals understand which open-source machine learning tools are best suited for building accurate and scalable binary classification models. It takes various datasets (up to 10 million rows and 1,000 features) and applies common algorithms like random forests, boosting, and neural networks using tools like H2O, xgboost, and scikit-learn. The output provides insights into training time, memory usage, and predictive accuracy (AUC) for each tool and dataset size. It's intended for data scientists, machine learning engineers, and analysts who need to choose the right tools for large-scale classification problems in business applications like credit scoring or fraud detection.

1,894 stars. No commits in the last 6 months.

Use this if you need to select an appropriate open-source machine learning library or framework for a new binary classification project, especially when working with medium to large datasets (millions of rows) and needing to balance speed, memory, and accuracy on a single machine.

Not ideal if your primary goal is interpretability of the models, if you need to compare different algorithms' theoretical foundations, or if you are focused on highly sparse data or use cases beyond binary classification.

data-science machine-learning-engineering binary-classification performance-benchmarking predictive-modeling
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 24 / 25

How are scores calculated?

Stars

1,894

Forks

330

Language

R

License

MIT

Last pushed

Sep 16, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/szilard/benchm-ml"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.