Synapses alternatives and similar libraries
Based on the "Machine Learning" category.
Alternatively, view Synapses alternatives based on common mentions on social networks and blogs.
Do you think we are missing an alternative of Synapses or a related project?
A lightweight library for neural networks that runs anywhere!
- Add one dependency to your project.
- Write a single import statement.
- Use a few pure functions.
You are all set!
It runs anywhere
It's compatible across languages
- The interface is common across languages.
It offers visualizations
Get an overview of a neural network by taking a brief look at its svg drawing.
You can specify the activation function and the weight distribution for the neurons of each layer. If this is not enough, edit the json instance of a network to be exactly what you have in mind.
The implementation is based on lazy list. The information flows smoothly. Everything is obtained at a single pass.
Data preprocessing is simple
By annotating the discrete and continuous attributes, you can create a preprocessor that encodes and decodes the datapoints.
Works for huge datasets
The functions that process big volumes of data, have an Iterable/Stream as argument. RAM should not get full!
It's well tested
Every function is tested for every language. Please take a look at the test projects.