A brief example of a neural network approximating a function of a known relationship. In this case the sine function.
This repository contains four files. Firstly a random dataset generator, this is just for the purposes of showing where the training and testing datasets came from. If you wish to try another function, such as a cosine function, simply change the line
y = math.sin(x)
to whatever you want in it's place.
Dependecies
To use this program you will need the follwing python packages:
- Keras
- Numpy
- Matplotlib
Usage
To run this file simply open your command line and navigate to your directory and run
python Model.py
The purpose of this repository is to illustrate a neural networks ability to approximate functions, feel Free to edit this code and have a play about!
NOTE
If you are wondering about the back end this example was built with tensorflow but will Run with Theano as well!