Soft Margin Rosenblatt's Perceptron
This is an adaption of the well-known Rosenblatt's perceptron. It uses a margin relaxation parameter in order to be more sensible to outliers. Also, There is a regularization parameter C which allows for solutions similar to Support Vector machines.
Drawback is that it is relatively sensitive to the parameters. One should not fiddle around with them too much.
If anybody interested in the theory, let me know. I would also ask you to inform me when you are using this code, what you are using it for and if you found a way to improve it.
Best, wedge