Welcome to the Javascript-RNN repository! This project is a Multi-Layer Recurrent Neural Network (RNN) implementation in JavaScript and HTML. The goal of this repository is to provide a simple and interactive way to understand the fundamentals of recurrent neural networks and sequence prediction.
This project aims to demonstrate the workings of a Multi-Layer Recurrent Neural Network (RNN), a type of neural network capable of processing sequential data and learning temporal dependencies. It is designed to be lightweight, browser-friendly, and accessible—making it an excellent learning tool for beginners and enthusiasts exploring sequence modeling.
- HTML: For the user interface.
- JavaScript: For the implementation of RNN algorithms, including recurrent connections, forward propagation, and backpropagation through time.
- Fully configurable multi-layer RNN with multiple hidden layers.
- Interactive network creation: set input size, hidden layer sizes, output size, and learning rate.
- Load training data via CSV upload or pasting sequences directly into the browser.
- Train the network in-browser with immediate feedback on predictions versus targets.
- Predict sequences dynamically and display results in a readable table.
- Save and load network models as JSON for reuse and experimentation.
- Lightweight, self-contained, and runs entirely in your browser—no backend required.
To get a local copy up and running, follow these simple steps:
Make sure you have a modern web browser installed (Chrome, Firefox, Edge, Safari).
- Clone the repository:
git clone https://github.com/matthewJamesAbbott/Javascript-RNN.git
- Open the
index.htmlfile in your web browser.