Description

A module for developing neural networks for machine learning models. This module also provides complex number arithmetic and matrix operations. To help you get started quickly, it also includes a built-in utility for loading and processing the MNIST dataset.

Constants

• activation — A default instance of class activation_function

Functions

• add()
• adjoint()
• argmax()
• argmin()
• calculate_error()
• complex_abs()
• complex_acos()
• complex_acosh()
• complex_add()
• complex_arg()
• complex_asin()
• complex_asinh()
• complex_atan()
• complex_atanh()
• complex_conjugate()
• complex_cos()
• complex_cosh()
• complex_division()
• complex_exp()
• complex_log()
• complex_log10()
• complex_multiply()
• complex_norm()
• complex_polar()
• complex_pow()
• complex_sin()
• complex_sinh()
• complex_sqrt()
• complex_substract()
• complex_tan()
• complex_tanh()
• conjugate()
• cross()
• determinant()
• dot()
• get_max_threads()
• identity()
• inverse()
• kronecker()
• matrix_random()
• max()
• mean()
• min()
• mnist_load_image()
• multiply()
• multiply_inplace()
• norm()
• normalise()
• ones()
• powm()
• prod()
• scalar_division()
• scalar_multiply()
• set_max_threads()
• sqrtm()
• squared_norm()
• substract()
• sum()
• trace()
• transpose()

Classes

• activation_function — This class provides activation functions for the neural network layers. Note: This class has no instantiation (constructor) method, please use constant activation instead
• complex
  
  

  Member Functions

  • construct()
  • imaginary()
  • print()
  • real()
  • string()
• matrix
  
  

  Member Functions

  • construct()
  • at()
  • print()
  • set()
  • to_array()
• mnist
  
  

  Member Functions

  • construct()
  • inputs()
  • outputs()
  • size()
• neural_network — A multi-layer neural network class. This class consists of input layer, hidden layers, and output layer
  
  

  Member Functions

  • construct() — The constructor. Create an instance of class neural_network
  • back_propagation() — Do a backward propagation process
  • feed_forward() — Move forward the input matrix through the model's layers
  • get_layer_biases()
  • get_layer_outputs()
  • get_layer_weights()
  • load()
  
  
  • And 6 more member functions

Example

import neural
 
start
    ' Construct the model with 784 inputs, 2 hidden layers, 10 outputs, and learning rate: 0.01
 
    model = neural_network([784, 20, 10, 10], 0.01)
 
 
    ' Load the MNIST dataset
 
    dataset_train = mnist("dataset/train-images.idx3-ubyte", "dataset/train-labels.idx1-ubyte")
    dataset_test = mnist("dataset/t10k-images.idx3-ubyte", "dataset/t10k-labels.idx1-ubyte")
 
 
    ' Train the model with epochs: 5
 
    train(model, dataset_train, 5)
 
 
    ' Test the model and calculate its accuracy
 
    test(model, dataset_test)
stop
 
function train(model, dataset, epochs = 3)
    writeln("Total training data: " & dataset.size())
 
    for i, 1, epochs
        writeln("Training... Epoch: " & i & " of " & epochs)
 
        for j, 0, dataset.size() - 1
            expected = dataset.outputs(j)
 
            model.feed_forward(dataset.inputs(j))
            model.back_propagation(expected)
 
            error = calculate_error(model.outputs(), expected)
        endfor
    endfor
stop
 
function test(model, dataset)
    total_errors = 0
    writeln("Testing")
 
    for i, 0, dataset.size() - 1
        input = dataset.inputs(i)
        expected = dataset.outputs(i)
 
        model.feed_forward(input)
        outputs = model.outputs()
 
        predicted = argmax(outputs)
        expc = argmax(expected)
        err = calculate_error(outputs, expected)
 
        if predicted != expc; total_errors += 1; endif
    endfor
 
    writeln("Total errors: " & total_errors & " - Total data: " & dataset.size() & " - Accuracy: " & (((dataset.size() - total_errors) / dataset.size()) * 100) & "%")
stop
 

Availability

This module is only available on Dinfio version 3.2.0 or later

Documentation version: 1.1.03. Updated on: 11 February 2026.
Dinfio is designed and written by Muhammad Faruq Nuruddinsyah. Copyright © 2014-2026. All rights reserved.