Skip to content

Latest commit

 

History

History
46 lines (32 loc) · 844 Bytes

README.md

File metadata and controls

46 lines (32 loc) · 844 Bytes

xgboost-js

A pure JavaScript implementation of XGBoost for both Node.js and browser environments.

Installation

npm install @fractal-solutions/xgboost-js

Usage

const { XGBoost } = require('@fractal-solutions/xgboost-js');

// Initialize the model
const model = new XGBoost({
    learningRate: 0.3,
    maxDepth: 4,
    minChildWeight: 1,
    numRounds: 100
});

// Train the model
model.fit(X_train, y_train);

// Make predictions
const predictions = model.predictBatch(X_test);

Features

  • Binary classification
  • Model serialization
  • Feature importance calculation
  • Comprehensive test suite
  • Pure JavaScript implementation
  • Works in both Node.js and browser environments

Documentation

For full documentation, see docs/xgboost.md

License

MIT © Fractal Solutions