AdamPerlinski

AdamPerlinski / micro-ml

Public

Tiny ML & statistics library for JS — 16 algorithms in ~56KB gzipped. Rust/WASM, zero dependencies.

33
1
100% credibility
Found Feb 13, 2026 at 18 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

micro-ml is a compact JavaScript library offering fast, lightweight machine learning tools for tasks like forecasting trends, smoothing data, classification, and clustering via WebAssembly.

How It Works

1
📊 Gather your numbers

You collect simple lists of data like monthly sales, sensor readings, or prices over time.

2
🔧 Add the tool

You easily include micro-ml in your project to analyze your data.

3
🎯 Choose your analysis

Pick what you need like forecasting ahead, smoothing wiggles, or grouping similar items.

4
✨ Run it and see magic

Your data transforms into clear trends, predictions, and insights in a flash.

5
📈 Explore the results

Review future forecasts, growth rates, or hidden patterns with confidence scores.

✅ Decide with confidence

You now plan smarter like stocking inventory or spotting issues early.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 33 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is micro-ml?

micro-ml delivers 16 machine learning and statistics algorithms to JavaScript in a 56KB gzipped package, powered by Rust and WebAssembly with zero dependencies. It fits trendlines to charts, forecasts sales, smooths sensor noise, clusters data, or classifies points—all sub-millisecond on typical datasets. Like github tiny c compiler for ML, it solves "I need quick predictions without 500KB bloat" for JS devs.

Why is it gaining traction?

At 56KB vs TensorFlow.js's 500KB+, it loads instantly in browsers while matching speeds via WASM—linear regression flies on 100k points. Live docs/demos plus npm simplicity hook devs; benchmarks beat pure JS libs like simple-statistics. Echoes tiny desk statistics vibe: maximal utility, minimal footprint.

Who should use this?

Frontend devs adding trend forecasts to dashboards, IoT engineers denoising streams, or analysts visualizing clusters in React/Vue apps. Ideal for micro ml tasks like sales prediction, anomaly peaks, or PCA down to 2D—no PhD required. Skip if scaling to millions of rows or needing neural nets.

Verdict

Strong pick for tiny ML in JS; excellent docs/benchmarks offset 1.0% credibility and 16 stars as an early project. Production-ready for prototypes/light use—watch for growth.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.