Skip to content

Files

Latest commit

 

History

History
9 lines (6 loc) · 753 Bytes

README.md

File metadata and controls

9 lines (6 loc) · 753 Bytes

Binary classification metrics preview

The notebook constists of an implementation of the stochastic gradient descent classifier for the problem based on a data set owned by Volker Lohweg (University of Applied Sciences, Ostwestfalen-Lippe) and available under this link.

The task is to distinguish real and forged banknotes based on certain features of their images. Trained model was tested using metrics described in Chapter 3 of Aurelien Geron, 'Hands - On Machine Learing with Scikit-Learn and TensorFlow', Helion SA, 2018.

scikit-learn library's implementations of machine learning methods and metrics were used. Matplotlib library was used for plotting and visualization.