NeuraFlow

A library to easily create neural networks in NodeJS

Installation

You can install NeuraFlow using npm:

npm install https://github.com/CaraPloof/neuraflow

Usage

NeuralNetwork class

The NeuralNetwork class allows you to create and train a neural network with customizable parameters:

• Constructor Parameters:
  • inputNodes: Number of input nodes.
  • hiddenNodes: Number of nodes in the hidden layer.
  • outputNodes: Number of output nodes.
  • learningRate: Learning rate for training.
  • epochs: Number of training epochs.
• Methods:
  • train(trainingInputs, trainingOutputs): Train the neural network using provided training data.
  • predict(input): Predict outputs for a given input after training.
  • export(path): Save the network's weights and biases to a specified file path.
  • import(path): Load the network's weights and biases from a specified file path.

Example

See the provided example in examples/xor.js to understand how to use NeuraFlow to solve the XOR problem.

Credits

Project initiated by DIDELOT Tim Aka Fufly / CaraPloof.

License

MIT