Polynomial regression
Louis Cialdella, louiscialdella@gmail.com
2013

This is a simple implementation of 2D polynomial regression using least squares, where the best order polynomial is selected using K-fold cross validation. 

The outside libraries being used are:
1. Numpy, for quickly finding the pseudoinverse and to use the argmin function.
2. Matplotlib, to graph the data.

A quick rundown of the files:
1. README: This one. 
2. Demo.py: The file which runs the demonstrations. By default, generates a random polynomial of order 1 to 10 and fits it as a demonstration. If the command line argument "nonpoly" is given, fits a sin wave instead. In both cases, 5-fold cross validation is used to select the order of the model polynomial.
3. Regression.py: Contains functions which allow generation, evaluation, and training of linear basis functions.
4. DataSets.py: Creates basic synthetic data and allows adding Gaussian noise.
5. GraphData.py: Uses matplotlib to plot the data.
6. CrossValidation.py: Compares the models using K-fold cross validation.