An incomplete list of Machine Learning frameworks or libraries written in JavaScript. All should run in browser context.
| Name | Description | License |
|---|---|---|
| ml.js | A compilation of Machine Learning tools in JavaScript | MIT |
| machine_learning | Implementations of multiple machine learning algorithms such as decision-tree | MIT |
| Name | Description | License |
|---|---|---|
| webdnn | DNN Running Framework that can be built with WebGPU, WebGL, or WebAssembly | MIT |
| synaptic.js | Generic neural network framework | MIT |
| deeplearning.js | Hardware-accelerated deep learning // machine learning // NumPy library for the web | Apache License V2 |
| kera.js | Run Keras models in the browser, with GPU support using WebGL | MIT |
| TensorFire | Framework for running neural networks in the browser, accelerated by WebGL. | To be open sourced |
| neurojs | deep learning framework focused on reinforcement learning | |
| deepforge | open-source visual development environment for deep learning | |
| neataptic | Blazing fast neuro-evolution & backpropagation for the browser | MIT |
| Feedforward Neural Networks | An implementation of feedforward neural networks based on wildml implementation | MIT |
| Kohonen Networks | An implementation of Kohonen Networks / self-organizing map (SOM) | MIT |
| ConvNetJS | [UNMAINTAINED] Cmmon neural network modules, including Convolutional Neural Network | MIT |
| Simple Feedforward Neural Networks | [UNMAINTAINED] Simple feed-forward neural network | MIT |
| Name | Description | License |
|---|---|---|
| gpu.js | Perform GPU-accelerated matrix computation with graceful fallback to JS | MIT |
| ml-matrix | Matrix manipulation and computation library created by mljs | MIT |
| math.js | Extensive math library compatible with the JS built-in math library | Apache License V2 |
| weblas | GPU Powered BLAS for Browsers | MIT |
| web-dsp | A client-side signal processing library built on WebAssembly | MIT |