mzusin / mz-ml

A collection of TypeScript-based machine learning helpers.

mz-ml - a collection of TypeScript-based machine learning helpers 🚀

This project is a collection of TypeScript machine learning helpers and utilities for the browser and Node.js environment. The modular approach allows to select only the required functions. It works well with all modern bundlers and supports tree shaking 🌲. The library is built using immutable/pure functions.

Linear Regression Features ⭐

• Train simple and multiple linear regression model with gradient descent.
• Evaluate model performance using statistical data (R-squared, mean squared error, pearson correlation coefficient).
• Batch size support.
• Shuffle support.
• Split data to training set, test set, and optional validation set.

Documentation 🔖

Main

• Linear Regression
• Simple Linear Regression
• Interfaces

Utilities

• Split Data

---

███    ███ ███████       ███    ███ ██      
████  ████    ███        ████  ████ ██      
██ ████ ██   ███   █████ ██ ████ ██ ██      
██  ██  ██  ███          ██  ██  ██ ██      
██      ██ ███████       ██      ██ ███████