The data set for this exercise comes from the UC Irvine Machine Learning data repository, which hosts a large collection of data sets for use with machine learning algorithms. The original data set for this exercise is available at https://archive.ics.uci.edu/ml/datasets/banknote+authentication#.
I have restructured the data set as a CSV file available here.
The original data set was prepared by taking pictures of both real and fake bank notes and then compressing those images via a wavelet transform. The researchers who prepared the data set then extracted four key features from each compressed input image, and then classified each instance as either a real note or a forged note.
Construct a classification model for this data set, and confirm that a simple linear classification model is not sufficient to successfully separate the two classes.
Next, construct a classification model using polynomial regression. What is the minimum degree polynomial model you need to successfully separate the two classes?
Finally, couple your polynomial model with Ridge regression, and demonstrate that it is still possible to successfully separate the two classes.
This assignment is due on Thursday, February 1.