Install the gem and add to the application's Gemfile by executing:

$ bundle add neuneu

If bundler is not being used to manage dependencies, install the gem by executing:

$ gem install neuneu

The first step is to create a dataset to handle our training examples. Here, we use the Memory implementation which wraps an array of examples.

Each of the training examples is composed of a vector (i.e. an array) of inputs and a vector of outputs. Here, each example consists in a value of temperature in Celsius degrees as the input, and a value in Fahrenheit degrees as the output:

require "neuneu"

data = [
  [[-40.0], [-40.0]],
  [[-10.0], [14.0]],
  [[0.0], [32.0]],
  [[8.0], [46.4]],
  [[15.0], [59.0]],
  [[22.0], [71.6]],
  [[38.0], [100.4]]
]

We normalize the inputs and outputs so that their values are shifted to the 0-1 range. We also tell Neuneu to shuffle values at each training epoch so that there's no influence of the order of the examples on the training process:

dataset = Neuneu::Dataset::Memory.new(data, transpose: true)
                                 .normalize!
                                 .shuffle!

Now we create a single-layer, single-neuron perceptron model, and train it on all examples during 50 epochs. We use the mean squared error (MSE) loss function and "leaky" rectified linear unit (LeakyReLU) as the transfer function:

model = Neuneu::Model.new.append(:input, 1)
                         .append(:dense, 1, transfer: :leaky_relu)
model.fit(dataset, epochs: 50, loss: :mean_squared_error)

We can show the evolution of the training loss across epochs in the terminal:

model.plot(width: 100)

Now that our model is trained we can make a prediction on novel/unknown input:

model.predict([[100]])
# => 211.7616137915528

After checking out the repo, run bin/setup to install dependencies. Then, run rake test to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.

To install this gem onto your local machine, run bundle exec rake install. To release a new version, update the version number in version.rb, and then run bundle exec rake release, which will create a git tag for the version, push git commits and the created tag, and push the .gem file to rubygems.org.

Bug reports and pull requests are welcome on GitHub at https://github.com/jefmathiot/neuneu.

The gem is available as open source under the terms of the MIT License.