Learnx

Learnx is a pure Elixir library for Machine Learning built on top of Nx for data and calculations and VegaLite for visualization.

The main idea is to provide established and widely used Machine Learning and Data Processing algorithms in a familiar manner to people coming from other languages, as well as providing comprehensive tools for newcomers.

A lot of focus is put in versatility, documentation and test-driven development. It is important to have precision en par with other implementations of these algorithms.

Scope of project

The project is intended to eventually to provide implementations for:

• Data Preprocessing
• Regression
  • [x] Linear
  • [x] Polynomial
  • [ ] SVR
  • [ ] NN
  • [ ] Naive Bayes
  • [ ] Decision Trees
  • [ ] And more...
• Classification
• Clustering
• Model Selection

Progress

Currently available functionality:

Regression

Simple Linear Regression

iex> x = [-1, 1, 3, 5]
iex> y = [6, 8, 10, 12]
iex> reg = Learnx.LinearRegression.fit(x, y)
iex> reg |> Learnx.LinearRegression.predict(2)
9.0
iex> reg |> Learnx.LinearRegression.predict([4, 6])
[11.0, 13.0]

Multiple Linear Regression

iex> x = [[1, 1], [1, 2], [2, 2], [2, 3]]
iex> y = [6, 8, 9, 11]
iex> reg = Learnx.LinearRegression.fit(x, y)
iex> reg.coef 
#Nx.Tensor<
  f32[2]
  [1.0, 2.0]
>
iex> reg.intercept
2.999993324279785
iex> reg |> Learnx.LinearRegression.predict([3, 5])
[15.999993324279785]
iex> reg |> Learnx.LinearRegression.predict([[3, 3], [3, 5]])
[11.999993324279785, 15.999993324279785]

Simple Polynomial Regression

iex> x = [-1, 1, 3, 5]    
iex> y = [6, 8, 9, 11]
iex> Learnx.PolynomialRegression.transform(x, 3)
#Nx.Tensor<
  s64[4][4]
  [
    [1, -1, 1, -1],
    [1, 1, 1, 1],
    [1, 3, 9, 27],
    [1, 5, 25, 125]
  ]
>
iex> reg = Learnx.PolynomialRegression.fit(x, y, 3) 
iex> reg.coef
#Nx.Tensor<
  f32[3]
  [0.9583308696746826, -0.24999012053012848, 0.04166489839553833]
>
iex> reg.intercept
#Nx.Tensor<
  f32
  7.249996185302734
>
iex> reg |> Learnx.PolynomialRegression.predict(0)
7.249996185302734
iex> reg |> Learnx.PolynomialRegression.predict([0,2])
[7.249996185302734, 8.500016689300537]

Installation

If available in Hex (not yet), the package can be installed by adding learnx to your list of dependencies in mix.exs:

def deps do
  [
    {:learnx, "~> 0.1.0"}
  ]
end

Documentation can be generated with ExDoc and published on HexDocs. Once published, the docs can be found at https://hexdocs.pm/learnx.

Independently started in 2022 by myself.