Wondering what’s next for npm?Check out our public roadmap! »

    @delight-im/neural-network

    1.0.1 • Public • Published

    JS-NeuralNetwork

    Neural networks in JavaScript. Well-documented and object-oriented.

    Installation

    • In the browser

      <script type="text/javascript" src="dist/browser-bundle.min.js"></script>
    • In Node.js

      $ npm install @delight-im/neural-network
      

      and

      var NeuralNetwork = require("@delight-im/neural-network");

    Usage

    • Creating a new instance

      • Neural network with 3 input neurons and 1 output neuron

        var network = new NeuralNetwork.Type.Feedforward(3, [], 1);
      • Neural network with 4 input neurons, 3 hidden neurons and 2 output neurons

        var network = new NeuralNetwork.Type.Feedforward(4, [ 3 ], 2);
      • Neural network with 6 input neurons, two hidden layers with 4 and 2 neurons, and 3 output neurons

        var network = new NeuralNetwork.Type.Feedforward(6, [ 4, 2 ], 3);
    • Passing any number of additional options to the network

      // pass an object containing the desired options as the fourth parameter
      var network = new NeuralNetwork.Type.Feedforward(3, [ 4 ], 1, {
          seed: 501935,
          learningRate: 0.3,
          hiddenLayerActivationFunction: new NeuralNetwork.Activation.HyperbolicTangent(),
          outputLayerActivationFunction: new NeuralNetwork.Activation.BinaryStep()
      });
    • Available activation functions

      new NeuralNetwork.Activation.ArcTangent();
      new NeuralNetwork.Activation.BinaryStep();
      new NeuralNetwork.Activation.GaussianFunction();
      new NeuralNetwork.Activation.HyperbolicTangent();
      new NeuralNetwork.Activation.Identity();
      new NeuralNetwork.Activation.LogisticFunction();
      new NeuralNetwork.Activation.RectifiedLinearUnit();
      new NeuralNetwork.Activation.RectifiedLinearUnit(0.01);
      new NeuralNetwork.Activation.SinusoidFunction();
    • Training the network using supervised batch ("all-at-once") learning

      // the first parameter is the array of inputs and the second parameter is the array of desired outputs
      // the third parameter is the optional number of iterations and the fourth parameter is the optional error threshold
      var error = network.trainBatch(
          [
              [0, 0, 1],
              [0, 1, 1],
              [1, 0, 1],
              [1, 1, 1]
          ],
          [
              [ 0 ],
              [ 1 ],
              [ 1 ],
              [ 0 ]
          ],
          60000,
          0.005
      );
    • Training the network using supervised online ("single-pattern") learning

      // the first parameter is the input and the second parameter is the desired output
      var error = network.train([0, 0, 1], [ 0 ]);
    • Asking the network to predict some output from a supplied input pattern

      // the single parameter is the input to process
      network.predict([ 0, 0, 1 ])
    • Saving the network with all its properties to a JSON string

      var jsonStr = JSON.stringify(network);
    • Restoring the network with all its properties from a JSON string

      var network = NeuralNetwork.Type.Feedforward.fromJson(jsonStr);

    Development

    • Prerequisites

      $ npm install -g uglify-js
      $ npm install -g browserify
      
    • Building the browser bundle

      $ browserify src/main.js --standalone NeuralNetwork > dist/browser-bundle.js
      $ uglifyjs dist/browser-bundle.js --compress --preamble "$(< src/header.js)" > dist/browser-bundle.min.js
      $ rm dist/browser-bundle.js
      
    • Running the Node.js examples

      $ node examples/node.js
      

    Contributing

    All contributions are welcome! If you wish to contribute, please create an issue first so that your feature, problem or question can be discussed.

    License

    Copyright (c) delight.im <info@delight.im>
    
    Licensed under the Apache License, Version 2.0 (the "License");
    you may not use this file except in compliance with the License.
    You may obtain a copy of the License at
    
      http://www.apache.org/licenses/LICENSE-2.0
    
    Unless required by applicable law or agreed to in writing, software
    distributed under the License is distributed on an "AS IS" BASIS,
    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    See the License for the specific language governing permissions and
    limitations under the License.
    

    Install

    npm i @delight-im/neural-network

    DownloadsWeekly Downloads

    2

    Version

    1.0.1

    License

    Apache-2.0

    Last publish

    Collaborators

    • avatar