@am/neuralnetwork - JSR

A simple neural network library for JavaScript and TypeScript.
@am/neuralnetwork is a simple neural network library written in TypeScript. It is designed to be easy to use and understand, while still being powerful enough to be useful in real-world applications.
Also, it's cross-runtime compatible, meaning you can use it in all Node.js, Deno, bun, and the browser. It
Constructing a neural network with @am/neuralnetwork involves a few key steps:
Here's a simple example of how to use @am/neuralnetwork to create a neural network that learns the XOR function. This example demonstrates the steps outlined above.
If you want more examples, check out the examples folder.
This package is feature complete for basic feedforward neural networks (Multi-Layer Perceptrons), but not feature complete as a general-purpose deep learning library.
The following features are not yet implemented:
Layer Types: Only Dense layers are available. Missing:
Evaluation Metrics: Only basic accuracy is implemented. Missing precision, recall, F1-score, confusion matrix, etc.
Data Processing: No built-in data augmentation, normalization utilities, or dataset handling
Advanced Features: No GPU acceleration, distributed training, transfer learning, or model visualization
Please provide a reason for reporting this package. We will review your report and take appropriate action.
Please review the JSR usage policy before submitting a report.
Import directly with a jsr specifier