https://github.com/kinoute/l-layers-xor-neural-network

A L-Layers XOR Neural Network using only Python and Numpy that learns to predict the XOR logic gates.
https://github.com/kinoute/l-layers-xor-neural-network

arguments deep-learning neural-network numpy python python3 script xor xor-neural-network

Last synced: about 1 month ago
JSON representation

A L-Layers XOR Neural Network using only Python and Numpy that learns to predict the XOR logic gates.

• Host: GitHub
• URL: https://github.com/kinoute/l-layers-xor-neural-network
• Owner: kinoute
• Created: 2019-06-16T18:08:47.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2019-06-24T13:35:17.000Z (over 5 years ago)
• Last Synced: 2024-11-09T14:45:09.358Z (3 months ago)
• Topics: arguments, deep-learning, neural-network, numpy, python, python3, script, xor, xor-neural-network
• Language: Python
• Size: 20.5 KB
• Stars: 2
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# L-Layers XOR Neural Network

A L-Layers XOR Neural Network using only Python and Numpy that learns to predict the XOR logic gates.

## Script

The script was initially made as a `.ipynb` jupyter file and later refactored as a class and a script where arguments can be passed to the neural network.

There are two files:

* `nn.py`: the main script that uses our NeuralNetwork class stored in the other file ;

* `nn_xor_class.py`: our NeuralNetwork class.

## Arguments

You can change the behavior of the Neural Network by using arguments when running the script. For example, you can change the activation function of the hidden layers, the learning rate etc. Here are the arguments allowed when running `nn.py`. All arguments are optional.

### Usage

```

python nn.py [-h] [-l LAYERS] [-u UNITS] [-s SIZE] [-i ITERATIONS] [-r LEARNING_RATE]

                    [-a {sigmoid,tanH,relu,leakyRelu}]

  -h, --help            show this help message and exit

  -l LAYERS, --layers LAYERS

                        Number of layers in your NN (including the output layer). Default: 4.

  -u UNITS, --units UNITS

                        Number of units in each hidden layer separated by a

                        comma (including output layer). Default:

                        4,2,1.

  -s SIZE, --size SIZE  How many examples should be generated in our training

                        set. Default: 1000.

  -i ITERATIONS, --iterations ITERATIONS

                        Choose the number of iterations we want. Default: 10000.

  -r LEARNING_RATE, --learning-rate LEARNING_RATE

                        Pick a Learning rate for your neural Network. Default: 1.5.

  -a {sigmoid,tanH,relu,leakyRelu}, --activation {sigmoid,tanH,relu,leakyRelu}

                        Activation function for your hidden layers. The output

                        layer will always be a sigmoid. Default: "tanH".

```